Skip to main content

Full text of "Palaeontology"

See other formats


JTION  NOIlfUllSNI  NVINOSH1IWS  S3ldVdan  LIBRARIES  SMITHSONIAN  INSTITUTION  NOIIDIIJ 
— co 


dal!  LIBRARIES  SMITHSONIAN  INSTITUTION  NOIlfUllSNI  NVINOSHlllNS  S31dVdan  LIBRAR 


rn  gey  ™ '$0"  rn  !!fj  rn  ™ rn  'sv°<y 

05  £ (/)  \ £ 05 

UTION  NOIlfUllSNI  NVIN0SH1IINS  S3ldVdan  LIBRARIES  SMITHSONIAN  INSTITUTION  NOlinil. 

O)  Z CO  Z ....  CO  z 


> 

ra  a l"|Z  LI  B RAR  I ES^SMITHSONIAN  INSTITUTION  NOIlfUllSNI  NVINOSHIIWS^SB  I d Vd  8 IT  LIBRAR 

— CO  ~ rn  — 'O 

(X. 

< 

__  o; 

O pcjx7  ^ q ^ O 

UTION  ~ NOIlfUllSNI  “nVINOSHIMS^S  3 IdVdaiT^LIBRARI  ES^  SMITHSONIAN^  INSTITUTION  NOIlfUL 
z i-  z ~ ~ - 


dail  LIBRARIES  SMITHSONIAN  INSTITUTION  NOIlfUllSNI  NVIN0SH1I1AIS  S3ldVdan  LIBRAR 

Z 05  Z V CO  Z CO  _ Z 


05  A'  Z ^ Z W5  2 CO 

JTION  NOIlfUllSNI  NVIN0SH1IWS  S3ldVdan  LIBRARIES  SMITHSONIAN  INSTITUTION  NOIlfUI. 

— - co 


9 2 2 W 2 3 %J|  3 2 


o -.,W  _ o ~ o 

z -J  Z _J  z 

dail  LIBRARIES  SMITHSONIAN  INSTITUTION  NOIlfUllSNI  NVIN0SH1MS  S3ldVdan  LIBRAR 

r r *7 


UTION  NOIlfUllSNI  NVIN0SH1IIAIS  S3ldVdan  LI  B RAR  I E S ^SMITHSONIAN  INSTITUTION ^ NOIlfUI. 

05  Z _ 05  z ....  05  z vv.  05 

s^!  '■#'!#=  I #1,1  1.^ 

.-4#  ?%  iO)i%  im 

'd  a llZ  LI  B RAR  I ES  SMITHSONIAN  INSTITUTION  NOIlfUllSNI  NVIN0SH1IINS°°S3  I d Vd  a IT  LIBRAR 
Z to  5 ^ ^ co  — „ <o  ^ _ z 


-j?  yj  2 ^ 2 (/) 

ioshiiwsv;S3 i a va a n libraries  Smithsonian  institution  NoiiniiisNi_NviNOSHiiws  S3ia\ 


fHSONIAN  INSTITUTION  NOlinillSNI  NVINOSH1MS  S3iavaai3  LIBRARIES  SMITHSONIAN  INSTIT 


josHiiiNS  S3iavaan  libraries  Smithsonian  institution  NoiinmsNi  nvinoshhins  S3ia\ 

to  2 to  2 v-  to  Z _ ^ 


^ ^ 2 'V  > 

rHSONIAN  INSTITUTION  NOlinillSNI  NVINOSHlIWS^'s  3 I a V a B 13  “ LI  B R AR  I E S SMITHSONIAN  _ INSTI1 

to  — go  . 2 <s>  z 

W X^^^x  “ UJ  XvWtT'x  7?, 


161  z!  x < (oft  36  z!  UF  < c Ufc 


gosHims  S3  lava  a n libraries  Smithsonian  institution  Nouniiism  nvinoshiiins  S3ia 


— X^titx^ 

K ' 

2 _ 

THSONIAN  INSTITUTION^NOlinillSNI- NVINOSHlI^s”  S3  I a Va  3 11  LI  B RAR  I ES  SMITHSONIAN  INSTI 

2 CO  2 <0  2 

2 ,/dfarr.  2 ± 


* W | | 'tj^  - 1 vW'  | W'  | 

NosHiiws^sa i ava a nZLi b rar i es^smithsonian  institution  NoiiniusNi_NviNosHiiws  sb  i a 


ITHSONIAN  INSTITUTION  NOlinillSNI  NVINOSH1IIAIS  S3iavaai3  LIBRARIES  SMITHSONIAN  INSTI 


m \ z 

NOSHIIINS”  S3  I aVaa  n~L!  BRAR  I ES^SMITHSONIAN'iNSTITUTIOn’  NOlinillSNI  NVINOSH1IIAIS  S3  I a 

in  z c/>  z ^ cn  z . ^ 


ITHSONIAN  INSTITUTION  NOlinillSNI  NVINOSH1IWS  S3IBVaail  LIBRARIES  SMITHSONIAN  _ I NSTI 


Palaeontology 

VOLUME  23  ■ PART  3 AUGUST  1980 


Published  by 

The  Palaeontological  Association  London 


Price  £1  5 


THE  PALAEONTOLOGICAL  ASSOCIATION 


The  Association  publishes  Palaeontology  and  Special  Papers  in  Palaeontology.  Details  of  membership  and  subscription  rates 
may  be  found  inside  the  back  cover. 

The  journal  Palaeontology  is  devoted  to  the  publication  of  papers  on  all  aspects  of  palaeontology.  Review  articles  are 
particularly  welcome,  and  short  papers  can  often  be  published  rapidly.  A high  standard  of  illustration  is  a feature  of  the 
journal.  Four  parts  are  published  each  year  and  are  sent  free  to  all  members  of  the  Association.  Typescripts  on  all-aspects  of 
palaeontology  and  stratigraphical  palaeontology  are  invited.  They  should  conform  in  style  to  those  already  published  in  this 
journal,  and  should  be  sent  to  Dr.  R.  A.  Fortey,  Palaeontological  Association,  Department  of  Palaeontology,  British  Museum 
(Natural  History),  Cromwell  Road,  London  SW7  5BD,  England,  who  will  supply  detailed  instructions  for  authors  on  request 
(these  were  published  in  Palaeontology  1977,  20,  pp.  921-929). 

Special  Papers  in  Palaeontology  is  a series  of  substantial  separate  works ; the  following  are  available  (post  free). 

1.  (for  1967):  Miospores  in  the  Coal  Seams  of  the  Carboniferous  of  Great  Britain,  by  A.  H.  v.  smith  and  m.  a. 
butterworth.  324  pp.,  72  text-figs.,  27  plates.  Price  £8  (U.S.  $19.50). 

2.  (for  1968):  Evolution  of  the  Shell  Structure  of  Articulate  Brachiopods,  by  a.  williams.  55  pp.,  27  text-figs.,  24  plates. 
Price  £5  (U.S.  $12). 

3.  (for  1968):  Upper  Maestrichtian  Radiolaria  of  California,  by  Helen  p.  foreman.  82  pp.,  8 plates.  Price  £3  (U.S.  $7.50). 

4.  (for  1969):  Lower  Turanian  Ammonites  from  Israel,  by  R.  freund  and  M.  raab.  83  pp.,  15  text-figs.,  10  plates.  Price  £3 
(U.S.  $7.50). 

5.  (for  1969):  Chitinozoa  from  the  OrdoVician  Viola  and  Fernvale  Limestones  of  the  Arbuckle  Mountains,  Oklahoma, 
by  w.  A.  M.  jenkins.  44  pp.,  10  text-figs.,  9 plates.  Price  £2  (U.S.  $5). 

6.  (for  1969):  Ammonoidea  from  the  Mata  Series  (Santonian-Maastrichtian)  of  New  Zealand,  by  r.  a.  henderson.  82  pp., 
13  text-figs.,  15  plates.  Price  £3  (U.S.  $7.50). 

7.  (for  1970):  Shell  Structure  of  the  Craniacea  and  other  Calcareous  Inarticulate  Brachiopoda,  by  a.  williams  and 
a.  d.  wright.  51  pp.,  17  text-figs.,  15  plates.  Price  £1-50  (U.S.  $4). 

8.  (for  1970):  Cenomanian  Ammonites  from  Southern  England,  by  w.  j.  Kennedy.  272  pp.,  5 tables,  64  plates.  Price  £8 
(U.S.  $19.50). 

9.  (for  1971):  Fish  from  the  Freshwater  Lower  Cretaceous  of  Victoria,  Australia,  with  Comments  on  the  Palaeo- 
environment,  by  M.  waldman.  130  pp.,  37  text-figs.,  18  plates.  Price  £5  (U.S.  $12). 

10.  (for  1971):  Upper  Cretaceous  Ostracoda  from  the  Carnarvon  Basin,  Western  Australia,  by  r.  h.  bate.  148  pp.,  43  text- 
figs.,  27  plates.  Price  £5  (U.S.  $12). 

11.  (for  1972):  Stromatolites  and  the  Biostratigraphy  of  the  Australian  Precambrian  and  Cambrian,  by  m.  r.  Walter. 
268  pp.,  55  text-figs.,  34  plates.  Price  £10  (U.S.  $24). 

12.  (for  1973):  Organisms  and  Continents  through  Time.  A Symposium  of  23  papers  edited  by  n.  f.  hughes.  340  pp., 
132  text-figs.  Price  £10  (U.S.  $24)  (published  with  the  Systematics  Association). 

13.  (for  1974):  Graptolite  studies  in  honour  of  O.  M.  B.  Bulman.  Edited  by  r.  b.  ricKards,  d.  e.  jackson,  and  c.  p.  hughes. 
261  pp.,  26  plates.  Price  £10  (U.S.  $24). 

14.  (for  1974):  Palaeogene  Foraminiferida  and  Palaeoecology,  Hampshire  and  Paris  Basins  and  the  English  Channel,  by 
j.  w.  Murray  and  c.  A.  wright.  171  pp.,  45  text-figs.,  20  plates.  Price  £8  (U.S.  $19.-50). 

15.  (for  1975):  Lower  and  Middle  Devonian  Conodonts  from  the  Broken  River  Embayment,  North  Queensland,  Australia, 
by  p.  G.  telford.  100  pp.,  9 text-figs.,  16  plates.  Price  £5-50  (U.S.  $13.50). 

16.  (for  1975):  The  Ostracod  Fauna  from  the  Santonian  Chalk  (Upper  Cretaceous)  of  Gingin,  Western  Australia,  by 
j.  w.  neale.  131  pp.,  40  text-figs.,  22  plates.  Price  £6-50  (U.S.  $16). 

17.  (for  1976):  Aspects  of  Ammonite  Biology,  Biogeography,  and  Biostratigraphy,  by  w.  j.  Kennedy  and  w.  a.  cobban. 
94  pp.,  24  text-figs.,  11  plates.  Price  £6  (U.S.  $14.50). 

18.  (for  1976):  Ostracoderm  Faunas  of  the  Delorme  and  Associated  Siluro-Devonian  Formations,  North  West  Territories, 
Canada,  by  d.  l.  dineley  and  e.  j.  loeffler.  218  pp.,  78  text-figs.,  33  plates.  Price  £20  (U.S.  $48). 

19.  (for  1977):  The  Palynology  of  Early  Tertiary  Sediments,  Ninetyeast  Ridge,  Indian  Ocean,  by  E.  m.  kemp  and  w.  k.  Harris. 
74  pp.,  2 text-figs.,  8 plates.  Price  £7  (U.S.  $17). 

20.  (for  1977):  Fossil  Priapulid  Worms,  by  s.  c.  morris.  159  pp.,  99  text-figs.,  30  plates.  Price  £16  (U.S.  $38.50). 

21.  (for  1978):  Devonian  Ammonoids  from  the  Appalachians  and  their  bearing  on  International  Zonation  and  Correla- 
tion, by  M.  R.  house.  70  pp.,  12  text-figs.,  10  plates.  Price  £12  (U.S.  $29). 

22.  (for  1978,  published  1979):  Curation  of  palaeontological  collections.  Ajoint  colloquium  of  The  Palaeontological  Associa- 
tion and  Geological  Curators’  Group.  Edited  by  m.  g.  bassett.  280  pp.,  53  text-figs.  Price  £25  (U.S.  $60). 

23.  (for  1979):  The  Devonian  System.  A Palaeontological  Association  International  Symposium.  Edited  by  M.  R.  house,  c.  t. 
scrutton,  and  m.  g.  bassett.  353  pp.,  102  text-figs.,  1 plate.  Price  £30  (U.S.  $72). 

24.  (for  1980):  Dinoflagellate  Cysts  and  Acritarchs  from  the  Eocene  of  southern  England,  by  j.  b.  bujak,  C.  downie,  g.  l. 
eaton,  and  G.  l.  williams.  104  pp.,  24  text-figs.,  22  plates.  Price  £15  (U.S.  $36). 

© The  Palaeontological  Association,  1980 


Cover:  Edriophus  levis  (Bather,  1914)  from  the  Middle  Ordovician  Trenton  Group  of  Kirkfield,  Ontario.  x2-5. 
Specimen  in  the  Smithsonian  Institution;  photograph  by  H.  B.  Whittington. 


DINOFLAGELLATE  CYSTS  FROM  THE 
UPPER  EOCENE-LOWER  OLIGOCENE  OF  THE 
ISLE  OF  WIGHT 

by  M.  LIENGJARERN,  L.  COSTA,  and  C.  DOWNIE 


Abstract.  The  Upper  Eocene  and  Oligocene  succession  of  the  Isle  of  Wight,  southern  England  (Headon  Beds 
to  Hamstead  Beds)  has  been  studied  palynologically.  Seventy-one  forms  of  dinoflagellate  cysts  are  recorded, 
including  two  new  genera,  Gerdiocysta  and  Vectidinium,  and  ten  new  species,  Distatodinium  scariosum, 
Eocladopyxis  tessellata,  G.  conopeum,  Glaphyrocysta  paupercula , Phelodinium  pachyceras,  P.  pumilum,  Phthano- 
peridinium  amiculum,  P.flebile,  Thalassiphora  fenestrata,  and  V.  stover i.  The  dinoflagellates  (with  the  exception 
of  Vectidinium)  are  marine  and  indicate  six  marine  incursions  or  partial  incursions  in  the  sequence;  the  mid- 
Headon  Beds,  the  Oyster  Bed  of  the  Bembridge  Marls,  the  Nematura  Band,  and  three  episodes  of  the  Upper 
Hamstead  Beds.  Correlation  with  the  Paris  Basin  indicates  that  the  base  of  the  Stampian  lies  near  the 
Nematura  Band. 

The  importance  of  dinoflagellate  cysts  in  the  stratigraphy  of  the  Palaeogene  has  been  emphasized 
in  several  recent  papers.  Many  long-standing  problems  in  the  Upper  Palaeocene  and  Lower  Eocene 
have  been  resolved  by  their  application,  but  problems  of  correlation  at  the  Eocene/Oligocene 
boundary  remain.  This  account  describes  the  dinoflagellate  cysts  from  the  classical  section  on  the  Isle 
of  Wight  in  southern  England.  The  initial  work  was  done  by  M.  Liengjarern  (1973)  and  has  been 
revised  recently  by  L.  Costa. 


STRATIGRAPHY 

The  sequences  in  the  Isle  of  Wight  span  the  Eocene/Oligocene  boundary,  and  the  placings  of  this 
boundary  have  varied  according  to  the  interpretation  of  different  authors  (see  Curry  et  al.  1978)  from 
the  base  of  the  Headon  Beds  to  the  base  of  the  Hamstead  Beds.  The  difficulties  in  correlation  and 
interpretation  are  largely  the  consequence  of  the  paralic  nature  of  the  deposits,  which  varied  from 
open-sea  to  freshwater  lacustrine  in  a complex  coastal  geography. 

Two  main  localities  are  reported  here.  In  the  east  of  the  island,  the  lower  part  of  the  succession, 
from  the  base  of  the  Lower  Headon  Beds  to  the  Bembridge  Marls,  is  exposed  continuously  in  the 
sea  cliffs  at  WhitecliffBay.  In  the  west,  the  upper  part  of  the  succession  (Bembridge  Marls-Hamstead 
Beds)  is  exposed  in  Bouldnor  and  Hamstead  cliffs  as  a continuous  sequence  (text-figs.  1 and  2). 

PALYNOLOGY 

All  the  samples  were  prepared  by  standard  palynological  methods.  Only  a few  samples  of  fluvial  sands  were 
barren,  the  remainder  yielded  rich  assemblages  of  palynomorphs,  including  pollen  and  spores,  plant  tissue, 
freshwater  algae,  dinoflagellate  cysts,  and  acritarchs.  Only  the  dinoflagellate  cysts  are  dealt  with  in  detail  in 
this  paper,  but  in  each  sample  the  proportions  of  pollen  and  spores,  Pediastrum,  dinoflagellates,  and  acritarchs 
based  on  counts  of  200  individuals  were  noted.  These  results  are  shown  in  Tables  1 and  2.  It  should  be  noted 
that  these  counts  were  made  after  sieving  through  a 20  ^m  sieve  and  that  consequently  pollen  is  under- 
represented. 

A complete  list  of  the  dinoflagellate  taxa  recorded  and  their  distribution  and  relative  abundances  are  shown 
in  Table  1 . Only  new  taxa  or  combinations,  or  taxa  necessitating  further  comment  are  described  here.  The  genera 
discussed  are  arranged  in  alphabetical  order;  suprageneric  dinoflagellate  cyst-taxa  are  not  employed  here. 
(Palaeontology,  Vol.  23,  Part  3,  1980,  pp.  475-499,  pis.  53-54.1 


476 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  1.  Stratigraphic  location  of  samples  collected  at  Hamstead  Cliff  (prefix  H)  and  at  Bouldnor  Cliff 

(prefix  B). 


The  terms  employed  in  the  descriptions  are  those  of  Williams  et  al.  (1973)  and  Evitt  et  al.  (1977).  In  some 
species,  the  arithmetical  mean  of  the  measurements  is  indicated  as  a figure  in  parenthesis.  The  reference  for 
holotypes  and  illustrated  specimens  is  given  with  reference  to  their  location  in  the  ‘England  Finder’  grid  system. 

Division  pyrrhophyta 
Class  DINOPHYCEAE  Fritsch  1935 
Order  peridiniales  Haeckel  1894 
Genus  distatodinium  Eaton  1976 

Type  species.  Distatodinium  craterum  Eaton  1976 


LIENGJARERN  ET  AL .:  EOCENE/OLIGOCENE  DINOFL AGELL ATES 


477 


Distatodinium  scariosum  sp.  nov. 
Plate  54,  fig.  3 


Name  derivation.  Latin,  scariosus,  thin,  papery. 

Diagnosis.  Distatodinium  with  broad,  hollow,  intratabular  processes  (usually  one  per  paraplate), 
oblate  to  subtriangular  in  cross-section,  distally  expanded,  and  bearing  a variable  number  of  thick 
secae  on  their  distal  margin.  Cingular  area  devoid  of  processes. 

Description.  The  central  body  ambitus  is  oval,  antero-posteriorly  elongate.  Apex  and  antapex  are  rounded; 
the  antapex  may  be  prolonged  into  a corona  formed  by  the  expanded  bases  of  the  antapical  processes. 

The  insertion  of  the  processes  on  the  central  body  is  subcircular,  oblate,  or  triangular.  The  processes  occur 
one  per  paraplate,  except  on  the  antapical  paraplate  (l'"'),  where  there  may  be  two  or  more  processes.  The 
degree  of  compression  of  the  processes  varies  on  a single  specimen;  some  processes  are  taeniate,  but  more 
commonly  they  are  oblate  to  subtriangular  and  are  open  distally.  The  distal  margin  of  the  processes  extends  into 
a variable  number  of  robust  secae,  sometimes  prolonged  into  fine  strands  which  might  connect  with  those  from 
near-by  processes. 

Two  of  the  apical  processes  are  considerably  smaller  than  the  other  two.  Cingular  and  sulcal  zones  are  free 
of  processes.  When  more  than  one  antapical  process  occurs,  their  proximal  sections  coalesce,  forming  a corona 
which  is  apparently  hollow. 


Holotype.  Slide  ML  1456,  R37/0,  sample  B 1 1 , Upper  Hamstead  Beds,  Lower  Oligocene,  Bouldnor  Cliff,  Isle 
of  Wight. 


Upper 

Headon 

Beds 


Middle 

Headon 

Beds 


Lower 

Headon 

Beds 


-WC24 


-WC21 


WC16 
WC15 


Brockenhurst  Bed 


Bembridge 
Marls 


Bembridge 

Limestone 


text-fig.  2.  Stratigraphic  location  of  samples  collected  at  Whitecliff  Bay  (prefix  WC). 


table  1.  Distribution  of  dinoflagellate  species 


478 


PALAEONTOLOGY,  VOLUME  23 


o § 


S S 


3 T* 


sia 

wa 

na 

88 

Z.8 

98 

WH 

iZH 

61H 

9H 

PH 

SSDAV 

SfDAV 

KDM 

SZDA\ 

PZD/A 

£ZDAY 

IZDAV 

OZDM 

6IDAV 

8IDAV 

Z.IDM 

9IDAV 

SIDAV 

HDAV 

UDAV 


l/~>  — i 


m 


XXX 

r— < v (N 


^vo  o 
r- 

— < vo  © 


?2 

si 


^ ^ ^ ^ ^ ^ 


s a £ ^ 

' §!- 
-5 

If' 


^ I 


a ^ 

s S a 


5 5 JS 


^ Cl,-0  ~ 

S'  Sf  s ? 


^ -a  ■*  3 


Q Q 


& 73  "a  v5  .5 


-s.-a  . : s s s 

qqqqqkjiiifens) 


s * Ps  . 


LIENGJARERN  ET  AL.\  EOCENE/OLIGOCENE  DINOFLAGELLATES 


<n  v r-  Tt 


x 


i/~>  m r- 


S g 
% -S 


i |-§ 


Its  I 
il  i 
1 H 
1 1I 

"a  -S  .2 
bo  5 as 


s s 

£ o 
R 60 


3 ‘5  ,® 
•3  33  ~ 
.a  <5  5 
^ .3 
- -2  h-K 
3 3 0’S 


tu  ® 

ft,^  § 

O a,  a. 


S3  1 

as  a 

v a -a 

ffS  is  C 


® 

3 S 


1 1 £ Si.  g I w 2 I S I 1 


®>§tj 

j .p  o'  a a ® <5  a 'C  S 's  be 
J ’as  ^ g 2 a .S  -5  -®  «*>  a s s?  js 

SaSSob-'Sb-S^SS' 

5,  "5  § § ft,  £ .g 


Q *g  s o fe  £ § *S5  Sd  2 -ft  ^ 

3 -3  I -§  ^gsl-S-S^oS 
" I 81 5 Is  S.g-1  ^ ^ bo® 
- -:  ^ -^co  Co  S hi  ^ S ^ ^ 


I I 
pi  ps 


tfcs* 


indicates  that  presence  is  below  1%. 


-E  1.  Distribution  of  dinoflagellate  < 


: —indicates  that  presence  is  below  1%. 


PLATE  53 


liengjarern  et  a!.,  Eocene/Oligocene  dinoflagellates 


LIENGJARERN  ET  AL EOCENE/OLIGOCENE  DINOFL AGELL ATES 


Measurements.  Holotype,  central  body  length  51  ju.m  including  operculum  (43  not  including  operculum), 
breadth  31  /urn;  process  length  5-15-5  /nm. 

Range.  Central  body  length  38-49  p.m  (not  including  operculum),  breadth  26-31  /urn;  process  length  5-16  y.m. 
Specimens  measured— 8. 

Comparisons.  The  broad,  usually  hollow  and  distally  open  processes,  commonly  unconnected 
distally,  distinguish  D.  scariosum  from  other  species  in  the  genus. 

Distribution.  Samples  Bll,  B15. 


Genus  emslandia  Gerlach  1961 
Type  species.  Emslandia  emslandensis  Gerlach  1961 

Emslandia  sp. 

Plate  54,  fig.  5 

Remarks.  This  species  of  Emslandia  has  a bulging  ventral  hypocyst  surface.  The  ambitus  is  sub- 
circular  to  ovoid.  The  epicyst  is  distally  rounded  and  is  prolonged  into  a very  short  apical  horn, 
subrectangular  in  outline,  with  distal  ending  truncate,  bifid  or  sometimes  produced  into  a variable 
number  of  short  solid  processes.  The  hypocyst  may  be  rounded  or  somewhat  pointed  medially 
(?compression)  and  sometimes  bears  a very  short,  solid  antapical  projection. 

The  autophragm  is  robust  but  does  not  exceed  2 ^m  in  thickness,  it  is  apparently  spongy,  perforate, 
and  its  outer  surface  is  scabrate.  Linear  thickenings  of  the  wall  appear  scattered  randomly  on  the 
autocyst;  sometimes  these  coalesce  on  portions  of  the  cyst  producing  irregular  reticulate  structures. 
Two  parallel  thickenings  of  the  autophragm  mark  the  cingular  margins. 

The  archeopyle  is  large,  type  P.  The  operculum  may  remain  attached  along  its  cingular  suture. 
Emslandia  sp.  differs  from  E.  emslandensis  by  its  thinner  autophragm  and  randomly  scattered 
ornament  of  linear  thickenings,  in  part  reticulate.  It  is  clearly  a distinct  species,  but  the  material  is  too 
badly  preserved  to  provide  satisfactory  types. 

Distribution.  Samples  WC  19-21,  23;  Middle  Headon  Beds,  Whitecliff  Bay,  Isle  of  Wight. 

Genus  eocladopyxis  Morgenroth  19666 
Type  species.  Eocladopyxis  peniculata  Morgenroth  19666 

Eocladopyxis  tessellata  sp.  nov. 

Plate  53,  fig.  6 

Name  derivation.  Latin,  tessellatum,  tessellated. 

Diagnosis.  Eocladopyxis  distinguished  by  abundant,  long,  solid,  intratabular  processes  which  end 
distally  in  fine  spines  repeatedly  furcated  and  reflexed.  The  central  body  is  moderately  compressed 
dorso-ventrally  and  its  ambitus  is  circular.  Archeopyle  type  A + 3A  + 6P.  Additional  sutures  may 
occur  randomly  between  any  pair  of  paraplates. 


EXPLANATION  OF  PLATE  53 

Fig.  1.  Gerdiocysta  conopeum  gen.  et  sp.  nov.,  SEM  showing  the  membrane  connecting  the  distal  ends  of  the 
processes,  x785. 

Fig.  2.  Gerdiocysta  conopeum  gen.  et  sp.  nov.,  holotype,  dorsal  view  showing  apical  archaeopyle,  x 500. 

Fig.  3.  Glaphyrocysta  paupercula  sp.  nov.,  holotype,  x 1000. 

Fig.  4.  Phthanoperidinium  amiculum  sp.  nov.,  holotype,  x 1000. 

Fig.  5.  Glaphyrocysta  paupercula  sp.  nov.,  specimen  with  reduced  processes,  x 1000. 

Fig.  6.  Eocladopyxis  tessellata  sp.  nov.,  holotype,  x 1000. 


482 


PALAEONTOLOGY,  VOLUME  23 


Description.  The  autocyst  is  moderately  to  strongly  compressed  dorso-ventrally  with  a circular  ambitus.  The 
autophragm  is  scabrate  and  is  produced  into  solid  intratabular  processes,  two  to  four,  sometimes  more,  per 
paraplate.  The  processes  are  only  slightly  flexible,  simple,  somewhat  expanded  proximally,  circular  in  cross- 
section;  distally  they  flare  into  a number  of  fine  spines  which  fork  repeatedly;  more  rarely  some  of  the  processes 
may  end  in  simple  bifurcations.  They  are  more  or  less  strongly  reflexed. 

The  archeopyle  appears  to  be  of  the  type  A + 3A  + 6P  although  it  is  possible  that  all  apical  plates  separate 
in  the  formation  of  the  archeopyle.  Additional  sutures  commonly  develop,  apparently  at  random,  between  any 
other  pair  of  paraplates,  both  on  the  epicyst  and  on  the  hypocyst. 

The  paratabulation  formula  may  sometimes  be  determined  on  the  basis  of  plate  separation,  and  is  4',  6",  6c, 
?5",  1 p.v.,  1 " ",  ?Xs.  Two  of  the  apical  paraplates  appear  to  be  larger  than  the  other  two.  The  precingular 
paraplates  are  of  roughly  the  same  size,  antero-posteriorly  elongate,  and  pentagonal  in  outline.  The  cingular 
paraplates  are  narrow  and  subrectangular,  and  frequently  bear  only  two  processes  each.  The  hypocyst  appears 
to  be  formed  by  five  large  postcingular  paraplates,  a prominent  posterior-ventral  paraplate  and  an  antapical 
paraplate,  but  these  are  only  rarely  evident  since  secondary  sutures  are  uncommon  on  the  hypocyst;  a number 
of  smaller  sulcal  paraplates  also  appear  to  be  present. 

Holotype.  Slide  ML  1451,  T51/2,  sample  WC25,  Middle  Headon  Beds,  Upper  Eocene,  Whitecliff  Bay,  Isle  of 
Wight. 

Measurements.  Holotype,  central  body  diameter,  37/xm;  process  length  8-15  /xm. 

Range.  Central  body  diameters  31-39  x 35-43  /xm;  process  length  4-5-10  /xm.  Specimens  measured— 11. 

Comparisons.  The  solid  processes,  paratabulation,  and  archeopyle  type  leave  no  doubt  as  to  the 
generic  allocation  of  E.  tessellata-,  however,  the  archeopyle  is  not  always  observable,  in  which  case 
the  specimens  closely  resemble  some  species  of  the  genus  Impletosphaeridium  Morgenroth  19666. 

E.  tessellata  differs  from  E.  peniculata  Morgenroth,  the  only  other  species  so  far  allocated  to  the 
genus,  in  its  larger  size  and  longer  processes.  The  process  terminations  in  E.  tessellata  are  more 
complex  than  in  E.  peniculata. 


Genus  gerdiocysta  gen.  nov. 

Name  derivation.  Latin,  gerdius,  weaver. 

Type  species.  Gerdiocysta  conopeum  sp.  nov. 

Diagnosis.  Cyst  ambitus  subcircular,  posteriorly  bilobed  or  rounded;  dorso-ventral  compression 
moderate  to  strong.  Pericyst  bearing  solid  penitabular  to  intratabular  processes  arranged  into 
annular,  soleate,  or  linear  complexes.  The  process  complexes  support  a reticulate  or  membraneous 
ectophragm,  which  on  the  dorsal  face  and  laterally  simulate  the  outline  of  the  paraplates.  On  the 
ventral  face,  a median  area  of  variable  size  is  free  of  ornament  and  ectophragm.  The  processes  on 
either  side  tend  to  be  linearly  oriented  more  or  less  parallel  to  the  ambitus;  the  ectophragm  on  the 
ventral  face  may  link  processes  from  different  paraplates. 

Inferred  tabulation  formula:  4',  6",  6c,  5'",  1 p.v.,  1"  ",  Os. 

Archeopyle  type  A,  with  zig-zag  margins  including  a slightly  offset  sulcal  notch.  Operculum  tetra- 
tabular,  commonly  free. 

Comments.  Gerdiocysta  is  similar  to  Areoligera  Lejeune-Carpentier  but  differs  strongly  in  the  posses- 
sion of  an  ectophragm,  which,  on  parts  or  all  of  the  dorsal  surface  of  the  cyst,  simulates  the  shape 
of  paraplates.  In  Areoligera  the  processes  may  be  joined  distally  or  laterally  by  trabeculae,  but 
these  are  sparse  and  are  loosely  interconnected  and  do  not  constitute  an  outer  reticulum  or 
membrane. 

The  genus  Riculacysta  Stover  1977,  resembles  Gerdiocysta  in  shape  and  in  possessing  a 
membranous  perforate  to  reticulate  ectophragm.  However,  in  Riculacysta  the  processes  are  not  in 
complexes,  and  are  restricted  to  the  ventro-lateral  and  lateral  zones  of  the  cyst.  The  ectophragm  on 
the  dorsal  surface  of  Riculacysta  lies  very  close  to  or  touches  the  autophragm  and  extends  across 
the  paraplate  sutures  in  that  region.  In  contrast  there  are  the  simulate  dorsal  complexes  in 
Gerdiocysta. 


LIENGJARERN  ET  AL EOCENE/OLIGOCENE  DINOFL AGELL ATES 


483 


Gerdiocysta  conopeum  sp.  nov. 

Plate  53,  figs.  1,  2 

Derivation  of  name.  Latin,  conopeum,  mosquito  net. 

Diagnosis.  Gerdiocysta  characterized  by  a finely  reticulate  to  membranous  perforate  simulate  ecto- 
phragm  developed  over  paraplates  V-4',  l"-5",  2"  '-4'",  1 p.v.,  and  \""\  an  arcuate  to  soleate 
complex  of  very  reduced  processes,  distally  free,  may  be  developed  on  paraplate  6".  The  process 
bases  are  connected  by  microgranular  thickenings  of  the  cyst  wall  which  form  low  ridges  within 
the  complexes;  these  thickenings  are  often  further  developed  into  an  intratabular  irregular,  coarse 
reticulum.  Individual  processes  are  solid,  slightly  fibrous,  and  distally  furcated.  The  median  ventral 
area  is  large. 

Description.  The  antapical  bilobation  of  the  central  body  may  be  moderately  or  only  weakly  marked.  The  dorsal 
convexity  and  ventral  depression  are  moderate.  The  endophragm  is  finely  granulose,  apparently  perforate.  The 
periphragm,  as  seen  on  the  process  walls  is  slightly  fibrous. 

The  process  complexes  are  determined  proximally  by  basal  granulose  thickenings  on  the  cyst  wall,  which 
form  a more  or  less  continuous  basal  ridge.  Distally,  the  simulate  ectophragm  is  well  developed  over  paraplates 
l'-4',  l"-5",  2"  '-4" ',  1 p.v.,  and  1"  ".  The  cingular  paraplates  2c-4c  may  bear  linear  complexes  of  processes 
which  may  or  may  not  be  distally  united.  A narrow  ectophragm  may  also  be  developed  on  the  ventral  sur- 
face, forming  an  arcuate  wing  bordering  the  central  area  free  of  ornament.  The  ectophragm  is  closely  perforate 
and  finely  reticulate  or  membranous;  both  types  may  combine  in  the  same  species. 

On  some  individuals,  the  processes  are  greatly  reduced,  no  ectophragm  is  developed,  but  a coarse  granulate 
basal  reticulum  extends  over  the  dorsal  plate  surfaces;  intermediate  forms  between  these  and  normal  specimens 
with  well-developed  processes  and  ectophragm  are  common. 

Holotype.  Slide  ML  1456,  E 29/2,  sample  Bll,  Upper  Hamstead  Beds,  Lower  Oligocene,  Bouldnor  Cliff,  Isle 
of  Wight. 

Measurements.  Holotype,  central  body  length  (operculum  not  included)  64  jim,  breadth  73  /urn,  processes  height 
up  to  20  /un. 

Range.  Central  body  length  (operculum  not  included)  47(54-7)64  fim,  breadth  63(68)79  pm,  process  length 
6-23  pm.  Specimens  measured— 15. 

Comparisons.  No  granulate  proximal  wall  thickenings  have  been  mentioned  in  the  description  of 
the  only  other  species  in  the  genus  G.  cassicula  (Drugg)  comb,  nov.,  which  also  appears  to  differ 
from  G.  conopeum  in  having  considerably  longer  processes  and  a more  prominently  bilobed 
antapex. 

Benedek  (1972,  pi.  1,  figs.  1 1 a-c)  illustrated  examples  as  Cyclonephelium  pastielsii  which  appear  to 
be  conspecific  with  G.  conopeum. 

Distribution.  Samples  B6,  7,  8,  11,  and  15.  Also  in  Lower  Lintforter  Beds  and  Ratinger  Beds  (early 
Rupelian),  Germany  and  Calcaire  de  Sannois  (early  Stampian),  France  (Chateauneuf,  pers.  comm.). 

Other  species  allocated  to  the  genus:  G.  cassicula  (Drugg)  comb.  nov.  = Areoligera  cassicula  Drugg  1 970,  p.  8 1 1 , 
figs.  2b,  3a-b. 


Genus  glaphyrocysta  Stover  and  Evitt  1978 
Type  species.  Glaphyrocysta  retiintexta  (Cookson  1975) 

Glaphyrocysta  pauper cula  sp.  nov. 

Plate  53,  figs.  3,  5 

Name  derivation.  Latin,  pauperculum,  diminutive  of  pauperculus,  poor. 

Diagnosis.  Central  body  compressed,  ambitus  subcircular  to  quadrangular,  with  or  without  antapical 
indentation.  Autophragm  microgranular,  finely  reticulate.  Processes  developed  along  a peripheral 


484 


PALAEONTOLOGY,  VOLUME  23 


band  of  varying  width,  leaving  relatively  prominent  mid-dorsal  and  mid-ventral  areas  free.  Processes 
solid,  fibrous,  simple  or  bifurcate.  The  processes  may  be  isolated  or  arranged  into  linear,  arcuate, 
soleate,  or  annular  complexes.  When  in  complexes  the  processes  are  joined  by  their  expanded 
proximal  parts;  a few  lateral  (rarely  distal)  trabeculae  may  occur.  The  complexes  have  a ragged 
appearance  distally.  Processes  from  different  complexes  may  be  joined  by  basal  ridge  and/or 
medially  by  sparse  trabeculae.  Processes  may  be  considerably  reduced  in  number  and  in  size. 

Processes  may  occur  on  some  or  all  of  the  paraplates  T- 4',  l"-5"  (rarely  on  6"),  1"  '-5"  ',  1 p.v., 
and  l"". 

The  archeopyle  is  apical  tetratabular,  type  A;  the  operculum  may  be  free  or  remain  attached.  The 
archeopyle  suture  has  a sulcal  notch  a little  offset  from  the  mid-body  line. 

Description.  The  central  body  is  moderately  to  strongly  compressed;  the  ambitus  varies  from  subcircular  to 
quadrangular,  the  antapex  is  rounded,  somewhat  indented  or  produced  into  one  or  two  unequal  lobes.  The 
autophragm  appears  microgranular  in  optical  section  and  is  finely  reticulate  in  surface  view. 

The  processes  are  variable  in  number,  size,  and  shape,  and  are  developed  along  an  ambital  line  of  variable 
width.  The  mid-dorsal  and  especially  the  mid-ventral  areas  are  free  of  ornament  and  relatively  prominent. 
Individual  processes,  when  well  developed,  are  solid,  slightly  fibrous  (most  noticeable  at  and  near  the  base), 
slender,  simple  or  bifurcate. 

The  processes  may  be  isolated,  although  some  alignment  may  often  be  evident,  or  arranged  into  complexes 
on  parts  of  the  cyst.  When  in  complexes,  the  processes  are  joined  proximally  by  low  ridges  formed  by  their 
expanded  bases;  sparse  ribbon-like  trabeculae  with  smooth  margins  may  also  occur  laterally,  and  only  rarely 
distally.  Processes  from  different  complexes  may  also  be  united  proximally  by  ridges  and  laterally  by  sparse 
trabeculae.  Process  complexes  are  normally  present  and  better  defined  on  the  apical,  dorsal  precingular,  and 
antapical  zones  of  the  cyst. 

All  apical  paraplates  bear  processes,  normally  arranged  into  four  or  three  annular  or  soleate  complexes; 
when  four,  two  are  smaller  and  tend  to  coalesce  into  a single  elliptical  complex.  Linear  to  arcuate  complexes 
may  occur  on  the  precingular  paraplates  l” -5"  (occasionally,  processes  occur  on  paraplate  6").  Towards  the 
periphery  of  the  dorsal  face  (2”  and  4")  the  complexes  may  be  soleate.  On  the  ventral  face,  linear  or 
somewhat  arcuate  complexes  may  be  clear  but  sometimes  the  peripheral  processes  may  coalesce  with  those 
from  postcingular  paraplates  and  become  part  of  a more  or  less  continuous  complex  parallel  to  the  ambitus. 
On  the  postcingular  paraplates  process  complexes  tend  to  lose  definition  and  to  form  a number  of  lines 
running  antero-posteriorly  near  the  periphery  of  both  dorsal  and  ventral  faces.  The  posterior  ventral  processes 
may  join  in  these  lines  or  be  separate  as  an  arcuate  complex.  A soleate  complex  is  frequently  observable 
on  paraplate  1" 

These  forms  with  more  or  less  well-defined  complexes  of  well-developed  processes  constitute  one  end  of  the 
range  of  variation  observed  in  this  species.  The  other  end  includes  forms  with  some  isolated  processes  reduced 
to  simple  spines  scattered  along  the  peripheral  and  dorsal  precingular  zones,  tending  to  form  two  to  four 
loosely  defined  lines  parallel  to  the  cyst  ambitus.  The  variability  between  both  extreme  types  is  continuous  in 
the  same  assemblage  and  cannot  be  applied  to  further  taxonomic  division. 

The  archeopyle  is  apical,  tetratabular;  the  opercula  may  be  free  or  may  remain  in  place.  A rather  shallow 
sulcal  notch,  relatively  little  offset  from  the  mid-cyst  line  is  observable  on  the  archeopyle  margin. 

Holotype.  Slide  ML  1455  P44/1,  sample  B8,  Upper  Hamstead  Beds,  Lower  Oligocene,  Bouldnor  Cliff,  Isle  of 
Wight. 


EXPLANATION  OF  PLATE  54 

Fig.  1.  Thalassiphora  fenestrata  sp.  nov.,  holotype,  dorsal  view,  showing  archaeopyle  and  fenestrations,  x 250. 
Fig.  2.  Phelodinium pumilum  sp.  nov.,  holotype,  dorsal  view  showing  archaeopyle  and  small  cavities  at  the  horns, 
x 1000. 

Fig.  3.  Distatodinium  scariosum  sp.  nov.,  holotype,  x 1000. 

Fig.  4.  Phelodinium  pachyceras  sp.  nov.,  holotype,  x 1000. 

Fig.  5.  Emslandia  sp.  Middle  Headon  Beds,  sample  WC20,  showing  precingular  archaeopyle  and  cingulum, 
x 500. 

Fig.  6.  Phthanoperidinium  flebile  sp.  nov.,  holotype,  x 1000. 

Fig.  7.  Vectidinium  stoveri  gen.  et  sp.  nov.,  holotype,  x 1000. 


^ V 


PLATE  54 


liengjarern  et  al.,  Eocene/Oligocene  dinoflagellates 


486 


PALAEONTOLOGY,  VOLUME  23 


Dimensions.  Holotype,  central  body  length  50  ^ m , breadth  59  ^m,  maximum  length  of  processes  10  ^m. 

Range.  Central  body  length  41(47-6)52  ^ m , breadth  48(57-4)64  ^m,  processes  length  (maximum)  6-20  ^m. 
Specimens  measured— 20. 

Comparison.  In  the  ragged  distal  appearance  of  the  ornament,  this  species  resembles  Glaphyrocysta 
divaricata  (Williams  and  Downie  1966),  but  no  process  complexes  are  defined  in  the  latter  where  the 
processes  are  united  distally  by  trabeculae  bearing  free  aculei  and/or  by  perforated  membranes  in  a 
more  complex  fashion  than  in  G.  paupercula. 

G.  paupercula  also  resembles  G.  intricata  (Eaton  1976),  G.  texta  (Bujak  1977),  and  G.  micro- 
fenestrata  (Bujak  1977),  where  individual  process  complexes  may  also  be  distinguished.  However, 
the  distal  connections  between  processes  in  those  species  are  always  more  complex  than  in 
G.  paupercula,  while  the  processes  are  rarely,  if  at  all,  united  distally.  G.  paupercula  may  be  a 
degenerate  offshoot  of  this  lineage. 

Genus  impletosphaeridium  Morgenroth  19666 
Type  species.  Impletosphaeridium  transfodum  Morgenroth  19666 

Impletosphaeridium  severinii  (Cookson  and  Cranwell  1967)  comb.  nov. 

1967  Baltisphaeridium  severinii  Cookson  and  Cranwell,  p.  208,  pi.  3,  figs.  1,  2. 

Comments.  This  species  is  transferred  to  Impletosphaeridium  in  view  of  its  solid  processes.  Some 
specimens  appear  to  show  archeopyle  sutures;  if  these  eventually  prove  to  be  consistent,  then 
I.  severinii  may  have  to  be  transferred  once  more  possibly  to  Eocladopyxis. 

Genus  phelodinium  Stover  and  Evitt  1978 
Type  species.  Phelodinium  pentagonale  (Corradini  1973)  Stover  and  Evitt  1978 

Phelodinium  pachyceras  sp.  nov. 

Plate  54,  fig.  4 

Name  derivation.  Greek,  pachys,  large,  keros,  horn. 

Diagnosis.  Phelodinium  characterized  by  apical  and  antapical  horns,  triangular  in  outline,  proximally 
broad,  and  distally  rounded.  Thin-walled  cysts  moderately  compressed  dorso-ventrally.  Endocyst 
sub-circular,  with  low  apical  and  antapical  lobes.  Apical  and  antapical  pericoels  well  developed; 
a narrow  ambital  pericoel  may  occur  between  the  horns. 

Pericyst  ornament  atabular  of  reduced  spinules.  Pericingulum  margins  indicated  by  folds  on  the 
periphragm.  Perisulcus  broad  and  shallow. 

Description.  The  cyst  is  thin-walled  and  usually  compressed  dorso-ventrally.  The  ambitus  has  convex  sides  and 
is  projected  into  three  prominent  horns;  these  are  triangular,  with  a broad  base  and  a blunt  distal  ending, 
and  are  subequal  in  size.  The  epipericyst  is  more  or  less  conical  and  somewhat  larger  than  the  hypopericyst; 
the  posterior  margin  of  the  hypopericyst  is  straight  or  slightly  concave. 

The  endocyst  is  rounded,  only  weakly  bilobed  posteriorly;  a rounded,  low  projection  into  the  base  of  the 
apical  horn  may  occur.  The  pericoels  are  well  developed  beneath  the  horns,  a narrow  pericoel  is  commonly 
present  between  the  antapical  horns.  The  ornament  is  reduced  to  small  spinules  or  granules,  apparently 
atabular  in  distribution.  Cingulum  relatively  wide,  not  indented;  its  margins  are  marked  by  two  parallel  folds 
on  the  periphragm.  The  sulcus  is  very  broad  posteriorly  but  narrows  markedly  towards  the  cingular  zone. 

The  archeopyle  is  difficult  to  observe  due  to  the  opercula  remaining  nearly  always  in  place,  but  the  wide 
posterior  archeopyle  suture  (H4),  lying  very  close  to  the  cingular  margin,  is  evident  on  most  specimens  observed. 

Holotype.  Slide  ML  1454,  H19/0,  sample  B6,  Upper  Hamstead  Beds,  Lower  Oligocene,  Bouldnor  Cliff,  Isle  of 
Wight. 


LIENGJ ARERN  ET  AL.:  EOCENE/OLIGOCENE  DINOFLAGELLATES 


487 


Dimensions.  Holotype,  pericyst  length  75  /am,  breadth  53  /xm,  endocyst  length  46  ^m,  breadth  53  /on,  apical 
horn  12  /un,  left  antapical  horn  15  /xm,  right  antapical  horn  13  /%m. 

Range.  Pericyst  length  57(65)77  /xm,  breadth  45(51-6)56  /xin,  apical  horn  6(9)12  /xm,  left  antapical  horn 
9(12)14  /xm,  right  antapical  horn  8(10)13  /xm.  Specimens  measured— 12. 

Distribution.  Upper  Hamstead  Beds  (B6,  B8),  ?Middle  Headon  Beds  WC19. 

Comparisons.  The  prominent  broad  horns  and  reduced  ornament,  as  well  as  a strong  dorso-ventral 
compression,  distinguish  P.  pachyceras  from  the  other  species  allocated  to  this  genus. 

Phelodinium  pumilum  sp.  nov. 

Plate  54,  fig.  2 

Name  derivation.  Latin,  pumilus,  dwarf. 

Diagnosis.  Phelodinium  of  small  size,  ambitus  bilaterally  asymmetrical  with  reduced  antapical  horns, 
right  antapical  broadly  rounded,  may  be  absent.  Apical  horn  small,  cylindrical  with  prominent  distal 
pore.  Pericingulum  relatively  wide,  marked  by  folds.  Sulcus  distinct. 

Description.  The  ambitus  varies  from  subcircular  to  distinctly  peridinioid:  the  bilateral  asymmetry  of  the  cyst 
is  nearly  always  evident.  The  dorso-ventral  compression  is  strong.  The  pericoels,  if  observable,  are  restricted  to 
the  cavities  beneath  the  horns.  The  cylindrical  apical  horn  is  distinctive,  its  truncated  distal  tip  bears  a 
prominent  pore  bordered  by  a thickening  of  the  periphragm.  The  left  antapical  horn  is  always  developed  and 
is  sharply  pointed  distally.  The  right  antapical  horn  is  often  absent  but  commonly  it  is  represented  by  a broad 
lobe. 

The  periphragm  is  very  thin  and  transparent  and  is  often  folded.  The  cingulum  is  only  very  slightly 
helicoid,  wide  in  relation  to  the  over-all  size  of  the  cyst;  anterior  and  posterior  cingular  sutures  are  indicated 
by  low  smooth  ridges  formed  by  folding  of  the  periphragm.  The  perisulcus  is  distinct. 

The  archeopyle  is  of  a type  and  shape  seen  in  species  of  Phelodinium.  Peri-  and  endoperculum  are  indistin- 
guishable. The  operculum  may  remain  attached  along  its  posterior  suture. 

Holotype.  Slide  ML  1450,  Q45/4,  sample  WC  23,  Middle  Headon  Beds,  Upper  Eocene,  Whitecliff  Bay,  Isle  of 
Wight. 

Dimensions.  Holotype,  pericyst  length  64  /xm,  breadth  54  /am,  apical  horn  6 /xm,  left  antapical  horn  5 /xm, 
right  antapical  horn  8 /xm. 

Range.  Pericyst  length  50(55)62  fim,  breadth  41(46-5)54  /xm,  apical  horn  3-5(4-5)6-4  (xm,  left  antapical  horn 
2-7(4-5)6-5  /xm,  right  antapical  horn  0(1)3  /xm.  Specimens  measured— 11. 

Comparisons.  The  small  size,  rounded  ambitus,  bilateral  asymmetry,  and  distinctive  apical  horn 
distinguish  this  species  from  all  known  Phelodinium  species.  Allocation  to  Phelodinium  is  based  on 
the  archeopyle  shape  and  relative  size,  the  absence  of  well-defined  pericoels  and  the  very  strong 
dorso-ventral  compression. 

Distribution.  Samples  WC18,  20,  21,  23,  and  25. 

Genus  phthanoperidinium  Drugg  and  Loeblich  1967 
Type  species.  Phthanoperidinium  amoenum  Drugg  and  Loeblich  1967. 

Phthanoperidinium  amiculum  sp.  nov. 

Plate  53,  fig.  4 

Name  derivation.  Latin,  amiculum , cloak. 

Diagnosis.  Phthanoperidinium  with  ambitus  rounded-pentagonal  to  suboval.  Epicyst  with  convex 
sides,  terminating  in  a short  apical  horn,  hypocyst  also  rounded,  produced  into  one,  very  occasion- 
ally two,  antapical  horns.  Peri-  and  endophragm  very  closely  appressed  except  beneath  the  horns. 


PALAEONTOLOGY,  VOLUME  23 


where  restricted  pericoels  develop.  Periphragm  ornamented  with  intratabular  spinules  and  peni- 
tabular  to  hyaline  sutural  ridges  with  smooth  to  slightly  denticulate  free  edges.  Laevigate  to  striate 
pandasutural  lines  may  be  distinct.  Pericingulum  and  perisulcus  laevigate,  bordered  by  membranes. 

Description.  The  pericyst  is  fusiform  in  lateral  view;  the  ambitus  is  rounded-peridinoid  to  subcircular  or  sub- 
oval. The  apical  horn  is  short,  trangular  and  distally  blunt.  The  left  antapical  horn  is  usually  well  developed. 
On  some  specimens,  a right  antapical  horn,  very  much  reduced,  may  occur;  on  most  specimens,  a projection 
of  the  sutural  ridges  takes  the  place  of  the  right  antapical  horn. 

The  intratabular  spines  are  small  and  solid,  distally  short  or  somewhat  capitate;  those  closer  to  the  paraplate 
periphery  may  be  arranged  in  a penitabular  ring.  The  ridges  are  hyaline  and  imperforate,  their  free  margins 
are  entire  or  very  slightly  serrate  to  denticulate;  the  height  of  the  ridges  normally  does  not  exceed  3 /j.m,  except 
along  the  cingular  sutures  where  they  may  be  up  to  5 ^m  in  height.  The  ridges  may  be  parasutural  or  peni- 
tabular in  position.  Narrow  laevigate  pandasutural  zones  are  normally  observable  on  parts  of  the  pericyst 
and,  on  some  specimens,  very  faint  striations,  perpendicular  to  the  margin  of  the  paraplate,  may  be  observable. 

The  paratabulation  formula  and  shape  of  the  paraplates  are  normal  for  the  genus.  The  pericingulum  is 
helicoid,  its  ends  being  offset  about  one  pericingular  width;  its  surface  is  laevigate.  The  perisulcus  is  relatively 
narrow,  moderately  excavated,  extending  anteriorly  to  nearly  a half  of  the  epicyst  height.  The  archeopyle  is 
formed  by  the  detachment  of  paraplate  2a,  but  it  is  only  rarely  observable.  Occasionally,  additional  sutures 
occur  along  the  margins  of  all  three  intercalary  plates. 

Holotype.  Slide  ML  1451,  K23/4,  sample  WC25,  Middle  Headon  Beds,  Upper  Eocene,  Whiteclilf  Bay,  Isle  of 
Wight. 

Dimensions.  Holotype,  pericyst  length  63  /im,  breadth  48  p.m,  apical  horn  7 /im,  left  antapical  horn  5-5  /un. 

Range.  Pericyst  length  47(55-5)63  ^m,  breadth  40(43)48  /urn,  apical  horn  3(5-5)7  /un,  left  antapical  horn 
3(5-5)7  p.m.  Specimens  measured — 10. 

Comparisons.  P.  eocenicum  (Cookson  and  Eisenack  1965)  appears  to  have  sutural  ridges  and  intra- 
tabular granules,  and  thus  resembles  P.  amiculum  in  the  style  of  ornament;  but  the  ambitus  in 
P.  eocenicum  is  fusiform  to  subpolygonal,  less  rounded  than  P.  amiculum  and  the  left  antapical 
horn  lies  closer  to  the  median  axis;  in  addition  both  intratabular  granules  and  sutural  ridges  are 
much  more  reduced  than  on  the  present  species. 

P.  alectrolophum  Eaton  1976  resembles  P.  amiculum  in  possessing  sutural-penitabular  ridges,  but 
these  bear  well-developed  spines  on  their  free  margins  and  the  intratabular  paraplate  surfaces  are 
smooth. 

Distribution.  Only  in  sample  WC25. 


Phthanoperidinium  flebile  sp.  nov. 

Plate  54,  fig.  6 

1978  Geiselodinium  cf.  geiseltalense  Krutzsch,  Chateauneuf  1978. 

Name  derivation.  Latin,  flebilis,  pathetic. 

Diagnosis.  Phthanoperidinium  with  ?partial  (not  continuous)  endophragm  occasionally  developed 
beneath  the  horns.  Ornament  intratabular  of  small  echinae  or  setae,  laevigate  sutural  bands  may  be 
observable.  Cingulum  indicated  by  a relatively  broad  equatorial  band  free  of  ornament. 

Description.  The  autocyst  ambitus  is  subcircular  to  oval,  but  is  frequently  folded  and  the  ambitus  may  appear 
somewhat  fusiform;  the  ambital  outline  is  little  affected  by  the  horns.  The  apical  horn  is  very  short,  sub- 
triangular  to  rectangular  in  outline;  its  apical  margin  may  be  smooth  or  may  bear  a tuft  of  short  spines,  to 
which  sometimes  the  entire  horn  is  reduced.  The  hypocyst  is  posteriorly  rounded,  and  may  bear  a very  short, 
sharp,  antapical  horn  slightly  to  the  left  of  the  median  line. 

The  autophragm  is  thin  and  bears  a variable  number  of  small  setae  or  echinae,  sometimes  reduced  to 
granules,  atabular  to  intratabular  in  distribution;  on  some  specimens  the  number  of  spines  is  reduced,  and 
these  may  adopt  a penitabular  arrangement.  Sutural  bands,  when  observable,  are  smooth  and  of  variable  width. 


LIENGJARERN  ET  AL.:  EOCENE/OLIGOCENE  DINOFL AGELL ATES 


489 


The  cingulum,  observable  on  some  specimens,  appears  as  a relatively  wide  band  free  of  ornament;  it  is  not 
indented.  The  sulcus  has  only  been  seen  on  one  specimen,  appearing  as  a very  broad,  slightly  depressed  area  with 
ornament  more  sparse  than  on  the  rest  of  the  ventral  autocyst  face. 

The  archeopyle,  rarely  observable,  is  intercalary  and  formed  by  the  loss  of  paraplate  2a;  additional  splitting 
may  sometimes  develop  along  the  lateral  sutures  of  paraplate  3",  but  only  very  rarely,  along  the  sutures  of  the 
remaining  paraplates  in  the  intercalary  series. 

Holotype.  Slide  ML  1453,  X27/3,  sample  H24,  Lower  Hamstead  Beds,  Lower  Oligocene,  Hamstead,  Isle  of 
Wight. 

Dimensions.  Holotype,  autocyst  length  39  /xm,  breadth  28  /xm,  apical  horn  5 ^m,  antapical  horn  1 /xm. 

Range.  Autocyst  length  31(35)42  /xm,  breadth  22(27)31  /am,  apical  horn  1(3)5  /xm,  antapical  horn  0(1)2  /xm. 
Specimens  measured— 20. 

Distribution.  Sample  H24;  Lower  Hamstead  Beds. 

Discussion.  P.  echinatum  most  closely  resembles  P.  flebile  in  its  ornament  of  spines,  but  in 
P.  echinatum  these  are  sutural  to  penitabular  (distribution  as  a single  simulate  ring),  whereas  they 
are  intratabular  to  atabular  in  P.  flebile. 

Occurrence.  Sample  H24,  and  at  base  of  Sannoisian  in  Paris  Basin  (Argile  Verte  de  Romainville). 


Genus  thalassiphora  Eisenack  and  Gocht  1960 
Type  species.  Thalassiphora  pelagica  (Eisenack  1938)  Eisenack  and  Gocht  1960 

Thalassiphora  fenestrata  sp.  nov. 

Plate  54,  fig.  1 

Name  derivation.  Latin,  fenestratus,  windowed. 

Diagnosis.  Thalassiphora  with  partial  fenestration  of  the  periphragm.  The  fenestration  is  restricted 
to  the  lateral  and  ventral  areas  of  the  periphragm.  The  extent  of  the  fenestrated  area  is  variable, 
but  it  never  extends  over  the  whole  dorsal  region.  The  perforations  are  large,  more  or  less  circular, 
and  may  be  closely  packed  forming  an  irregular  reticulum.  The  ventral  flange  of  the  pericyst  is  narrow 
and  is  fenestrated  throughout. 

Description.  This  species  is  similar  to  T.  pelagica  in  shape  and  in  wall  structure  but  the  extension  of  the 
periphragm  on  the  ventral  side  appears  to  be  more  reduced  than  is  common  in  T.  pelagica,  that  is,  the 
ventral  lacuna  is  larger.  Perforations  develop  in  the  periphragm  in  ventral  and  lateral  areas  and  disappear 
towards  the  mid-dorsal  area.  Between  these  perforations,  the  fibres  are  more  loosely  packed.  A large  number 
of  smaller  perforations  occur  between  the  larger  fenestrations,  the  latter  are  of  variable  diameter  tending  to  be 
larger  closer  to  the  ambitus.  Ventrally,  the  pericyst  occurs  as  a relatively  narrow  flange  which  is  strongly 
fenestrate  throughout.  The  antapical  keel  may  often  be  reduced  or,  sometimes,  absent. 

Holotype.  Slide  ML  1449  U16/2,  sample  WC14,  Middle  Headon  Beds,  Whitecliff  Bay,  Isle  of  Wight. 
Measurements.  Holotype,  endocyst  81  x 67  /xm,  pericyst  diameter  150  /xm. 

Range.  Endocyst  73(77)89  x 59(67)77  /xm,  pericyst  diameter  126(154)182  /xm.  Specimens  measured— 10. 

Comments.  This  species,  which  is  apparently  restricted  in  distribution  to  the  latest  Eocene  and  ?early 
Oligocene,  seems  to  be  an  intermediate  form  between  T.  reticulata  Morgenroth  1966a,  which  is 
characteristic  of  younger  Oligocene  deposits  and  whose  pericoel  is  fenestrate  virtually  all  over,  and 
T.  pelagica. 

Distribution.  Samples  WC 13-23. 


490 


PALAEONTOLOGY,  VOLUME  23 


Genus  vectidinium  gen.  nov. 

Name  derivation.  Latin,  Vectis,  Roman  name  for  the  Isle  of  Wight. 

Type  species.  Vectidinium  stoveri  sp.  nov. 

Diagnosis.  Single-walled  proximate  peridinioid  cysts,  moderately  compressed  dorso-ventrally, 
ambitus  subpentagonal  or  subcircular  to  oval  or  somewhat  fusiform.  Epicyst  and  hypocyst  of 
approximately  equal  size.  Epicyst  may  or  may  not  extend  into  a short  apical  horn;  apical  pore  always 
present.  Hypocyst  semicircular  or  bilobed;  left  antapical  horn  present  or  absent,  right  antapical  horn 
commonly  present. 

Autophragm  with  atabular  or  intratabular  to  penitabular  ornament  of  small  granules,  spinules  or 
baculae,  which  may  be  reduced  in  size  and/or  number.  Narrow  laevigate  pandasutural  zones  may  be 
observable.  Paratabulation  formula,  when  determinable,  4',  3a,  1",  Oc,  5",  2'"',  Os.  When 
observable  paraplate  1"  is  rhombic,  antero-posteriorly  elongate,  and  relatively  large. 

Cingulum  and  sulcus  distinct.  The  cingulum  is  wide  relative  to  over-all  autocyst  size,  not  indented, 
non-  or  moderately  helicoid.  Sulcus  shallow  and  broad  on  the  hypocyst.  Archeopyle  combination 
type  31  3P  3"-5",  accessory  sutures  may  occur  along  cingular  margin  of  the  remaining  precingular 
paraplates.  Opercula  free. 

Comparisons.  Vectidinium  differs  from  Palaeoperidinium  Deflandre  1934,  and  from  Saeptodinium 
Harris  1975,  in  that  the  apical  paraplate  3'  is  not  included  in  the  archeopyle.  From  Saeptodinium  it 
also  differs  in  being  single  walled  and  usually  having  intratabular  or  penitabular  ornament.  From 
Palaeoperidinium  it  differs  in  the  presence  of  ornament  and  its  much  smaller  size. 

Ginginodinium  Cookson  and  Eisenack  1960,  Laciniadinium  McIntyre  1975,  and  Lunatodinium 
Brideaux  and  McIntyre  1973,  all  have  a 31  3P  3"-5"  archeopyle,  and  they  also  resemble  Vectidinium 
in  the  type  of  ornament.  Ginginodinium  is  double  walled,  and  in  the  formation  of  the  archeopyle  the 
three  dorsal  precingular  paraplates  (3"-5")  always  remain  attached  along  their  cingular  margins 
(Lentin  and  Williams  1975,  p.  95).  Laciniadinium  has  a single  opercular  piece  31  3P  3"-5"  which 
always  remains  attached  to  the  cyst  along  its  posterior  margin,  like  a flap.  In  Vectidinium  whenever 
the  archeopyle  is  present,  the  operculum  is  detached  and  some  doubt  remains  as  to  whether  this  is 
simple  or  compound.  Lunatodinium  (a  Lower  Cretaceous  genus)  was  described  as  having  an  archeo- 
pyle formed  by  the  loss  of  the  three  dorsal  precingular  paraplates.  However,  Lentin  and  Williams 
(1975,  pp.  96  and  116)  included  this  genus  in  the  pericysts,  possessing  a 31  3P  archeopyle.  This 
appears  to  be  so  from  the  original  illustration  of  Lunatodinium  (Brideaux  and  McIntyre  1973, 
figs.  1-13).  The  genus  is  stated  to  have  a circular  or  subcircular  outline. 

Cysts  of  the  Recent  freshwater  dinoflagellate  Peridinium  resemble  Vectidinium  in  the  type  and 
distribution  of  the  ornament,  but  they  are  normally  cavate  and  the  archeopyle  is  formed  by  the 
detachment  of  plates  along  a transapical  suture,  type  A3I3P. 

Vectidinium  stoveri  sp.  nov. 

Plate  54,  fig.  7 

Name  derivation.  This  species  has  been  named  after  Lew  Stover. 

Diagnosis.  As  for  the  genus. 

Description.  The  dorso-ventral  compression  of  these  cysts  is  normally  slight,  and  some  specimens  may  be 
oriented  in  apical  or  antapical  view;  in  lateral  view  the  cysts  are  somewhat  fusiform  or  oval.  The  epicyst  has 
strongly  convex  sides  which  may  merge  imperceptibly  in  a very  short,  blunt  apical  horn  with  a solid  tip  on 
which  sits  a pore;  the  apical  horn  may  be  absent,  and  the  epicyst  apex  is  then  invaginate.  The  hypocyst  is  com- 
monly broadly  rounded  posteriorly,  but  some  specimens  may  show  a weak  bilobation  on  the  antapex.  The 
short,  eccentrically  located  left  antapical  horn  may  be  present  or  absent. 

The  ornament  varies  in  density  and  shape.  When  the  ornament  is  baculate  or  of  short  processes  their  distal 
endings  are  often  T-shaped  and  may  be  linked  to  those  from  near-by  processes,  giving  the  appearance  of  a 


LIENGJ ARERN  ET  AL EOCENE/OLIGOCENE  DINOFL AGELLATES 


491 


tectum  supported  by  columellae  in  optical  section;  sometimes  the  ornament  is  very  reduced  in  size  and  mostly 
consisting  of  granules.  The  ornament  may  be  densely  or  sparsely  arranged  on  the  paraplate  surface,  the  most 
peripheral  elements  tending  to  be  arranged  along  simulate  rings.  Laevigate  pandasutural  zones,  usually  narrow, 
are  present  but  are  not  always  clearly  visible. 

Cingulum  and  sulcus  are  distinct,  both  being  marked  by  low  ridges  or  folds  on  the  autophragm.  The 
cingulum  is  relatively  wide,  slightly  helicoid  or  circular,  not  indented;  intratabular  ornament  and  smooth 
pandasutural  zones  may  be  observable  on  the  cingular  surface,  but  the  number  of  cingular  paraplates  has  not 
been  determined  with  certainty.  The  sulcus  is  also  broad  and  shallow,  and  extends  approximately  half-way  to 
the  apex.  The  shape  and  relative  size  of  individual  paraplates  are  difficult  to  determine  because  of  very  small 
size  and  transparent  autophragm  of  these  cysts. 

When  present,  the  archeopyle  is  formed  by  complete  detachment  of  plates  la-3a,  3"-5".  On  some  specimens, 
accessory  archeopyle  sutures  develop  along  most  of  the  anterior  margin  of  the  cingulum,  but  both  portions  of 
the  cyst  usually  remain  attached  along  a narrow  band,  presumably  corresponding  to  the  sulcus.  The  operculum 
is  always  free,  but  it  has  not  been  possible  to  determine  whether  this  is  formed  by  a single  piece  or  is 
compound,  since  isolated  opercula  have  not  been  observed— a fact  suggesting  that  the  operculum  may  be 
compound,  disintegrating  into  the  very  small  individual  paraplates  which  would  easily  be  lost  in  sieving  of  the 
organic  residue  during  preparation. 

Holotype.  Slide  ML  1452,  U43/3,  sample  WC34,  Upper  Headon  Beds,  Upper  Eocene,  Whitecliff  Bay,  Isle  of 
Wight. 

Measurements.  Holotype,  autocyst  length  37  ^m,  breadth  42  ^m,  apical  horn  1 /un,  left  antapical  horn  1 ^m, 
width  of  cingulum  4 ^m. 

Range.  Autocyst  length  30(35-5)41  ^m,  breadth  24-5(31)42  ^m,  apical  horn  0(2)4-2  /urn,  left  antapical  horn 
0(l)4-5  /xm,  width  of  cingulum  2-7(3-6)4  ^m.  Specimens  measured — 24. 

Distribution.  The  distribution  of  Vectidinium  stoveri  in  the  section  studied  deserves  some  special 
attention  since  it  constitutes  monospecific  assemblages  at  some  horizons,  and  has  not  been  found  in 
association  with  any  other  dinoflagellate  cysts.  These  horizons  yield  ostracod  assemblages  of  type  III 
(Keen  1972,  1977);  these  have  been  stated  by  Keen  to  indicate  brackish-water  conditions  (salinity 
3-9%).  V.  stoveri  is  thought  to  be  a non-marine  dinoflagellate  cyst,  and  possibly  a good  indicator 
of  oligohaline  conditions;  it  is  recorded  from  samples  WC34,  35,  and  HI 9. 


PALYNOLOGICAL  ASSEMBLAGES  AND  DEPOSITIONAL  ENVIRONMENTS 

The  Upper  Eocene-Lower  Oligocene  of  the  Isle  of  Wight  was  deposited  under  widely  variable 
environmental  conditions.  The  area  of  deposition  has  been  likened  to  an  embayment,  limited  to  the 
north  and  south  by  the  Portsdown  and  the  Sandown-Brixton  anticlines  respectively,  and  opening 
towards  the  sea  to  the  east  and  south-east.  At  times  this  sea  penetrated  into  the  basin.  At  other 
times  an  eastward  flowing  river  system  occupied  the  area  (Keen  1977).  The  conditions  ranged  from 
shallow,  near-shore  open  sea,  to  brackish-water  lagoons— with  or  without  connection  to  the  sea— 
to  freshwater  lacustrine  or  fluviatile  environments.  These  changes  are  reflected  in  the  palyno- 
assemblages,  and  are  especially  noticeable  in  the  relative  proportions  of  different  classes  of  palyno- 
morphs  as  well  as  in  the  composition  of  the  microplankton  assemblages  where  these  occur. 

Palaeoecological  studies  of  palynomorph  assemblages  and  particularly  of  dinoflagellate  cysts  are 
currently  in  their  preliminary  stages,  and  no  work  on  the  palaeoenvironmental  interpretation  of 
Tertiary  palyno-assemblages  from  paralic  areas  has  yet  been  published.  However,  the  assemblages 
recovered  here  may  be  correlated  to  particular  environmental  conditions  by  using,  as  a control,  the 
existing  information  on  the  distribution  of  dinocysts  in  Tertiary  to  Recent  sediments,  as  well  as  the 
sedimentological  and  faunal  evidence  available  from  the  sections  studied.  The  foraminifera 
(Murray  and  Wright  1974),  molluscs  (Daley  1973),  and  ostracods  (Keen  1972,  1977;  Haskins  1969) 
from  the  Upper  Eocene-Lower  Oligocene  sections  of  the  Isle  of  Wight  have  yielded  a considerable 
volume  of  data  that  can  be  used  in  assessing  the  meaning  of  the  palynological  assemblages  recovered. 


492 


PALAEONTOLOGY,  VOLUME  23 


The  major  components  of  the  palynological  assemblages  are  indicated  in  Table  2.  They  clearly 
fall  into  two  groups,  one  with  marine  dinoflagellates  present;  the  other  non-marine  samples  contain 
only  terrigenous  freshwater  or  lagoonal  elements. 

The  non-marine  group  shows  considerable  variation,  particularly  in  the  proportions  of  Pediastrum 
Meyen,  which  may  contribute  from  0 to  over  90%  of  the  assemblage.  In  some  samples  there  is  also  a 
considerable  contribution  from  non-marine  dinoflagellates.  These  non-marine  samples  are  asso- 
ciated with  various  lithologies  ranging  from  limestone  through  to  sands  and  no  particular  pattern 
has  so  far  been  determined.  It  is  evident,  particularly  from  the  work  of  Keen,  that  the  salinities  vary 
from  fresh  to  oligohaline  water.  The  environments  of  deposition  include  evidently  freshwater 
lacustrine,  fluvial,  flood-plain,  and  bay-head  situations.  The  control  over  the  relative  abundance  of 
Pediastrum  Meyen  is  not  understood.  It  is  notably  more  common  in  the  Bembridge  Marls  in  the 
west  of  the  island.  In  marine  sediments  it  is  present  only  in  very  small  numbers  and  is  probably 
allochthonous.  It  is  most  abundant  in  situations  that  could  be  interpreted  as  oligohaline  water. 


table  2.  General  character  of  palynological  assemblages.  P & S— pollen  and  spores;  Ped —Pediastrum  spp.; 
MD- marine  dinoflagellates;  fd— freshwater  dinoflagellates;  ‘r’  indicates  that  dinoflagellates  are  all  reworked 

from  older  strata. 


Sample  % P & S % Ped  % MD  % fd 
Whitecliff  Bay 

Bembridge  Marls 


WC67 

99 

— 

l(r) 

— 

WC66 

99 

1 

x (r) 

— 

WC65 

94 

2 

4(r) 

— 

WC64 

99 

1 

x(r) 

— 

WC63 

91 

8 

l(r) 

— 

WC62 

74 

26 

— 

— 

WC61 

100 

— 

— 

— 

WC60 

100 

— 

— 

WC59 

97 

3 

— 

— 

WC58 

100 

_ 

— 

WC56 

99 

1 

— 

WC55 

13 

80 

7 

— 

WC55A 

16 

84 

X 

— 

WC54 

100 

— 

— 

— 

WC53 

100 

— 

— 

— 

Bembridge  Limestone 
WC51  100 

WC49 

100 

— 

— 

— 

Osborne  Beds 
WC47 

100 

WC46 

58 

42 

— 

— 

WC45 

36 

64 

— 

— 

WC44 

100 

— 

— 

— 

WC43 

100 

— 

— 

— 

WC42 

10 

90 

— 

— 

WC41 

97 

3 

— 

— 

Upper  Headon  Beds 
WC40  22 

78 

WC39 

86 

14 

— 

WC38 

97 

3 

— 

WC37 

93 

6 

— 

1 

Sample 

% P & s 

% Ped 

% MD 

% fd 

WC36 

99 

1 



x 

WC35 

62 

1 

— 

37 

WC34 

73 

14 

— 

13 

WC33 

94 

6 

— 

_ 

WC31 

62 

38 

— 

— 

WC30 

100 

— 

— 

— 

WC29 

61 

39 

— 

— 

WC28 

90 

10 

— 

— 

WC26  100 

Middle  Headon  Beds 

— 

— 

— 

WC25 

74 

— 

26 

— 

WC24 

84 

16 

— 

WC23 

54 

46 

— 

WC21 

58 

2 

40 

— 

WC20 

40 

— 

60 

— 

WC19 

31 

— 

69 

— 

WC18 

28 

— 

72 

— 

WC17 

38 

— 

62 

— 

WC16 

29 

— 

71 

— 

WC15 

25 

75 

— 

WC14 

20 

80 

— 

WC13  37 

Lower  Headon  Beds 

X 

37 

WC12 

72 

28 

— 

— 

WC9 

80 

20 

— 

— 

WC8 

45 

55 

— 

— 

WC7 

25 

75 

— 

— 

WC6 

100 

— 

— 

— 

WC5 

7 

93 

— 

— 

WC4 

80 

20 

— 

— 

WC3 

96 

4 

— 

— 

WC2 

100 

— 

— 

— 

WC1 

98 

2 

— 

— 

LIENGJ ARERN  ET  AL .:  EOCENE/OLIGOCENE  DINOFL AGELLATES 


493 


Sample  % P & S % Ped  % MD  % fd 


Hamstead  Cliff 
Lower  Hamstead  Beds 


H36 

54 

46 

— 

— 

H35 

55 

45 

— 

— 

H34 

88 

12 

_ 

— 

H33 

82 

18 

— 

— 

H32 

75 

25 

— 

— 

H31 

63 

37 

— 

— 

H30 

81 

19 

— 

— 

H29 

41 

59 

— 

— 

H28 

64 

36 

— 

— 

H27 

97 

3 

— 

— 

H26 

96 

4 

— 

— 

H25 

96 

3 

— 

— 

H24 

94 

1 

6 

— 

H23 

42 

13 

45 

— 

H22 

84 

16 

- 

— 

H21 

87 

13 

x 

— 

H20 

78 

22 

— 

— 

H19 

63 

29 

— 

8 

H18 

76 

24 

— 

— 

H17  73 

Bembridge  Marls 

27 

— 

— 

H16 

76 

24 

— 

— 

H15 

34 

66 

— 

— 

H14 

67 

33 

— 

— 

H13 

64 

36 

— 

— 

H12 

36 

64 

— 

— 

H10 

34 

66 

— 

X 

Sample 

% P & S 

% Ped 

% md 

% fd 

H9 

57 

43 

_ 

x 

H8 

75 

25 

_ 

x 

H7 

73 

27 

_ 

_ 

H6 

35 

6 

59 

_ 

H4 

49 

X 

51 

_ 

H3  44 

Bembridge  Limestone 

56 

H2 

16 

84 

_ 

— 

HI  100 

Bouldnor  Cliff 
Upper  Hamstead  Beds 

B15 

92 

— 

8 

_ 

B14 

94 

— 

6 

_ 

B13 

96 

4 

_ 

_ 

B12 

100 

— 

_ 

_ 

B 1 1 

24 

2 

74 

_ 

BIO 

75 

25 

— 

_ 

B9 

42 

58 

_ 

_ 

B8 

8 

3 

89 

— 

B7 

28 

2 

70 

_ 

B6 

52 

29 

19 

_ 

B5  63 

Lower  Hamstead  Beds 

37 

x 

— 

B4 

100 

_ 

_ 

_ 

B3 

100 

_ 

_ 

_ 

B2 

100 

— 

— 

— 

B1 

52 

48 

— 

_ 

Non-marine  dinoflagellates  are  represented  by  a single  species,  Vectidinium  stoveri  which  is  present 
only  in  three  samples,  WC34,  35,  and  H19.  It  is  associated  with  ostracod  assemblage  III  of  Keen, 
indicating  brackish-water  conditions. 

Marine  samples  are  characterized  by  the  presence  of  marine  dinoflagellate  cysts  and  acritarchs. 
They  can  be  classified  into  a number  of  types  according  to  their  diversity  and  the  dominant  species. 
Since  these  types  occur  in  stratigraphic  order  and  are  associated  with  a series  of  marine  incursions  it  is 
convenient  to  discuss  them  in  stratigraphic  sequence. 

The  Middle  Headon  Beds  transgression 
Four  assemblage  types  are  present: 

Assemblage  1.  The  Brockenhurst  Bed  and  Psammobia  Beds  (samples  WC13-21)  are  characterized  by 
assemblages  with  forty  or  more  species  of  dinoflagellate  cysts  dominated  by  Homotryblium  plectilum 
which  makes  up  30-70%  of  the  microplankton;  other  abundant  species  are  Spiniferites  ramosus, 
Adnatosphaeridium  reticulense,  and  Phthanoperidinium  cometum.  These  assemblages  are  associated 
with  ostracod  assemblage  type  VI  and  indicate  open-sea  conditions,  the  major  transgressive  episode 
in  the  sequence  studied. 

Assemblages  2-4.  The  succeeding  Venus  Bed  contains  three  different  assemblage  types  showing  a 
marked  reduction  in  the  number  of  species  present  and  in  their  relative  abundance. 

Type  2,  occurring  in  sample  WC23,  has  less  than  thirty  species  and  is  dominated  by  H.  pallidum 
and  P.  cometum,  the  latter  a species  evidently  tolerant  of  reduced  salinities  in  estuarine  or  lagoonal 
environments. 


494 


PALAEONTOLOGY,  VOLUME  23 


Type  3,  occurring  in  sample  WC24,  has  only  seventeen  species  and  is  dominated  by  broken  species 
of  H.  plectilum  associated  in  assemblage  1 with  open-sea  conditions.  Here  these  are  thought  to  be 
allochthonous.  H.  pallidum  is  the  next  most  common  species. 

Type  4,  occurring  in  sample  WC25,  is  dominated  by  Eocladopyxis  tessellata  and  P.  cometum. 

These  three  assemblages  appear  to  indicate  a period  of  regression  with  restriction  of  marine  access 
to  the  area.  Keen  refers  the  ostracod  assemblages  in  these  beds  to  his  type  V,  indicating  salinities  in 
the  range  of  16-5-33%. 

The  Lower  Bembridge  Marl  transgression 

Assemblage  types  5-7  are  associated  with  the  Oyster  Bed. 

In  the  east,  sample  WC55  yielded  assemblage  type  5,  where  dinoflagellates  made  up  only  7%  of  the 
palynomorphs.  No  clearly  dominant  species  was  present,  the  commonest  being  Chiropteridium 
aspinatum,  Glaphyrocysta  microfenestrata,  Homotryblium  pallidum,  and  Paralecaniella  indentata. 

In  the  west,  assemblage  type  6 is  monospecific;  Phthanoperidinium  levimurale  makes  up  51%  of  the 
palynomorphs  in  sample  H4.  Assemblage  type  7 is  also  monospecific,  G.  microfenestrata  making  up 
59%  of  the  palynomorphs  in  sample  H6. 

The  significance  of  these  three  diverse  assemblages  from  the  Oyster  Bed  is  made  clearer  by 
consideration  of  the  fauna.  Molluscs,  foraminifera,  and  ostracods  all  indicate  brackish  estuarine 
conditions.  Assemblage  type  5 is  associated  with  Keen’s  type  V indicating  near-marine  conditions; 
the  assemblages  from  the  west,  however,  are  associated  with  his  type  IV,  indicating  lower  salinities 
(9-16%).  This  seems  to  mean  that  the  monospecific  assemblages  with  P.  levimurale  and  G.  micro- 
fenestrata are  composed  of  more  or  less  stenohaline  species,  since  both  also  occur  in  open  marine 
conditions.  They  appear  to  have  flourished  in  this  estuarine  situation  since  they  are  particularly 
abundant,  more  so  than  any  of  the  species  in  the  east,  where  the  assemblage,  although  poorer  in 
relative  numbers,  has  a greater  variety  of  marine  species  and,  although  still  estuarine,  appears  to  have 
better  connection  with  the  open  sea. 

The  Lower  Hamstead  Bed  transgression 

Assemblage  types  8 and  9 are  associated  with  a marine  incursion  at  the  horizon  of  the  Nematura 
Band. 

Assemblage  type  8,  sample  H23,  contains  only  four  species  and  is  dominated  by  Adnatosphaeridium 
reticulense.  Only  13%  of  the  palynomorphs  are  dinoflagellates.  Assemblage  type  9,  an  even  poorer 
assemblage  from  H24  immediately  above,  is  on  the  other  hand  dominated  by  P.  flebile.  Ostracods 
from  the  Nematura  Band  show  the  presence  of  assemblage  type  IV  characteristic  of  mesohaline 
conditions. 

The  Upper  Hamstead  Bed  transgressions 

Six  different  dinoflagellate  assemblages  (types  10-15)  have  been  found  in  the  Upper  Hamstead  Beds 
and  the  palynology  appears  to  show  the  presence  of  three  different  invasions  of  saline  water. 

The  first  incursion  corresponds  to  the  Cerithium  Bed  and  contains  assemblage  types  10-12. 
Assemblage  type  10,  sample  B6,  contains  19%  dinoflagellates  with  only  a few  species  represented  and 
is  dominated  by  G.  microfenestratum  and  P.  cometum,  both  of  which,  although  known  from  other 
marine  sediments,  have  previously  been  noted  in  assemblage  types  7 and  4 and  2,  with  reduced 
salinities  associated  with  Keen’s  types  IV  and  V.  Keen  (1972)  finds  that  the  Cerithium  Bed  also  yields 
assemblages  of  types  IV  and  V.  Assemblage  type  1 1 in  sample  B7  is  also  impoverished  in  species,  but 
is  dominated  by  small  acritarchs  of  the  Micrhystridium  group,  which  accounts  for  about  60%  of  the 
palynomorphs.  Assemblage  type  12,  sample  B8,  is  more  varied  and  richer  in  numbers,  but  G.  pauper- 
cula  accounts  for  most  of  these. 

Taken  together  these  three  samples  indicate  a marine  influence,  which,  however,  did  not  achieve 
fully  marine  conditions  in  this  locality,  the  area  remaining  meso-  to  polyhaline. 

The  second  incursion  is  represented  only  by  assemblage  type  13,  sample  Bll.  That  it  is  a 
separate  episode  is  indicated  by  the  intervention  of  samples  B9  and  10  which  contain  only  terrigenous 


LIENGJARERN  ET  AL..  EOCENE/OLIGOCENE  DINOFL AGELL ATES 


495 


pollen  and  spores  and  the  ?freshwater  alga  Pediastrum.  Assemblage  type  13  appears  to  represent 
more  fully  marine  conditions  with  many  new  species  appearing.  The  dominant  species  is  H.  pallidum, 
which  also  dominates  in  assemblage  type  2 ( Venus  Bed)  and  is  abundant  in  type  5 (Oyster  Bed, 
WhiteclifF  Bay).  Here  it  is  associated  with  Gerdiocysta  conopeum.  The  conditions  indicated  are  still 
not  yet  fully  marine,  but  must  closely  approach  that  condition. 

The  third  incursion  is  represented  by  assemblage  types  14  (sample  B14)  and  15  (sample  B15).  That 
this  is  a separate  episode  is  indicated  by  the  intervention  of  the  purely  terrigenous  palynological 
assemblages  in  samples  B12  and  13.  The  second  and  third  incursions  together  form  the  Corbula  Bed. 
Assemblage  type  14  is  a poor  monospecific  one  comprising  only  Phthanoperidinium  cometum.  It 
probably  indicates  low  salinities.  Type  15,  however,  is  somewhat  richer  and  is  particularly  so  in  the 
variety  and  lack  of  any  clearly  dominant  species.  Micrhystridium , Lejeunia  tenella,  Hystrichokolpoma 
salacium,  and  P.  amoenum  are  prominent,  the  last  three  being  known  only  from  open  marine 
sediments.  It  is  believed  that  these  two  samples  B14  and  B15  represent  the  beginning  of  a major 
transgression,  the  culmination  of  which  is  not  represented  due  to  erosion  of  the  succeeding  beds. 


DINOFLAGELLATE  CYST  STRATIGRAPHY 

The  distribution  of  dinoflagellates  is  shown  in  Table  1. 

The  first  dinoflagellate  assemblages  appear  in  the  Brockenhurst  Bed  associated  with  the  Middle 
Headon  transgression.  Detailed  comparison  between  the  dinoflagellate  assemblages  from  the  Solent 
Formation  and  the  marine  sediments  of  the  underlying  Barton  Formation  is  not  possible  at  present, 
since  little  information  on  the  dinoflagellate  content  of  the  Barton  Beds  has  so  far  been  published 
(Bujak  1976).  However,  from  unpublished  evidence  (Bujak  1973),  it  appears  that,  notwithstanding 
the  intervening  regression  represented  by  the  Becton  and  Lower  Headon  Beds,  only  minor  changes 
take  place  in  the  composition  of  the  assemblages  between  the  uppermost  marine  beds  of  the  Barton 
Formation  and  the  lower  part  of  the  Solent  Formation  (Middle  Headon  Beds).  The  number  of  species 
that  first  appear  in  the  Middle  Headon  Beds  is  very  small,  but  they  include  Rhombodinium  perforatum 
and  Thalassiphora  fenestrata,  and  the  possibility  remains  that  some  of  these  may  also  occur  in  the 
Barton  Beds;  the  number  of  apparent  extinctions  is  also  limited,  and  their  stratigraphic  significance, 
which  may  be  only  local,  cannot  be  assessed  at  this  stage. 

As  the  assemblages  become  impoverished  towards  the  upper  part  of  the  Middle  Headon  Beds, 
among  the  dinoflagellate  species  disappearing  from  the  assemblages  are  Areosphaeridium  diktyo- 
plokus,  Cordosphaeridium  funiculatum,  Distatodinium  ellipticum,  Palaeocystodinium  golzowense, 
R.  draco,  R.  perforatum,  and  T.  velata. 

Other  taxa,  Emslandia  sp.,  Eocladopyxis  tessellata,  and  Phelodinium  pumilum,  make  their  first 
appearance  in  the  section  here.  These  species  first  appearing  within  the  upper  part  of  the  Middle 
Headon  Beds  are  all  new  and  so  their  stratigraphic  value,  if  any,  cannot  be  stated. 

The  Bembridge  transgression,  represented  by  the  Oyster  Bed,  yields  poorly  diversified  assem- 
blages. These,  in  terms  of  their  species  content,  show  a somewhat  closer  relationship  to  the  Middle 
Headon  Beds  than  to  the  Upper  Hamstead  Beds.  The  Bembridge  Oyster  Bed  at  Whitecliff  Bay 
registers  the  last  known  occurrence  in  England  of  Chiropteridium  aspinatum,  Impletosphaeridium 
severinii,  Homotryblium  oceanicum,  and  Leptodinium  incompositum. 

The  Lower  Hamsted  Bed  transgression,  represented  by  a thin  sequence  including  the  Nematura 
Bed,  also  provides  a poor  assemblage  consisting  mainly  of  long-ranging  species.  One  species, 
Phthanoperidinium  flebile  is,  however,  apparently  confined  to  this  horizon. 

A very  pronounced  break  in  the  dinocyst  succession  is  evident  in  the  final  transgressions  of  the 
Upper  Hamstead  Beds.  Out  of  a total  of  sixty-eight  dinoflagellate  species  recorded,  only  nineteen 
are  common  to  the  Solent  and  Hamstead  Formations;  thirty-four  species  disappear  below  the  base 
of  the  Hamstead  Beds,  and  fifteen  species  are  first  recorded  within  the  latter.  The  marked  renewal 
of  the  assemblages  registered  between  the  two  main  marine  episodes  in  the  sequence  is  to  some  extent 
environmentally  controlled,  since  some  of  the  species  missing  in  the  Headon  Bed  are  known  to  persist 
elsewhere  into  the  Oligocene,  such  as  C.  aspinatum,  Cordosphaeridium  cantharellum,  D.  ellipticum. 


496 


PALAEONTOLOGY,  VOLUME  23 


Hystrichokolpoma  rigaudiae,  Kisselovia  coleothrypta,  R.  draco,  T.  velata,  and  T.  pelagica.  Two 
species,  however,  which  fail  to  reappear  are  R.  perforatum  and  A.  diktyoplokus,  whose  absence  seems 
to  be  stratigraphically  important. 

A number  of  species  make  their  first  appearance  here  and  some  of  them  are  thought  to  be 
stratigraphically  important.  These  are  Gerdiocysta  conopeum,  Heteraulacacysta  cf.  companula, 
Phthanoperidium  amoenum,  Wetzeliella  gochtii,  and  W.  symmetrica  incisa.  Other  appearances  of 
possible  significance  are  Phelodinium  pachyceras  and  D.  scariosum. 


CORRELATION  WITH  OTHER  EUROPEAN  AREAS 

Paris  Basin 

Curry  et  al.  (1978)  correlate  the  Middle  Headon  Beds  with  part  of  the  Marnes  a Pholadomya 
ludensis,  i.e.  with  the  deposits  of  the  Ludian  transgression  of  the  Paris  Basin.  Both  formations  yield 
rich  dinoflagellate  assemblages.  A description  of  those  from  France  has  been  given  by  Chateauneuf 
(1978).  Most  of  the  species  recorded  by  him  are  present  in  the  Middle  Headon  Beds  but  there  is 
none  of  sufficiently  restricted  range  to  allow  confident  correlation  on  the  basis  of  the  dinoflagellates, 
except  that  R.  perforatum  (which  appears  for  the  first  time  in  the  mid-Headon  Beds  in  England) 
also  appears  for  the  first  time  in  small  numbers  in  the  top  Marinesian  and  more  commonly  in  the 
Ludian.  R.  perforatum,  previously  mentioned  from  the  Barton  Beds  (Costa  and  Downie  1976)  is  in 
fact  a separate  species  (Bujak,  in  press).  A marked  distinction  between  the  Ludian  assemblages  and 
those  from  the  Headon  Beds  is  the  remarkable  abundance  of  H.  plectilum  in  the  Isle  of  Wight  and 
its  apparent  absence  from  the  Ludian. 

The  impoverished  assemblages  from  the  Bembridge  Oyster  Bed  yield  little  of  correlative  value,  but 
the  abundance  of  C.  aspinatum  does  correspond  with  the  prominence  of  this  species  in  assemblages 
from  the  Ludian  Marnes  a Lucines  (Chateauneuf  1978). 

The  equally  poor  assemblages  from  the  vicinity  of  the  Nematura  Band  do,  however,  show  some 
marked  similarities  to  those  of  the  Argile  Verte  de  Romainville  at  the  base  of  the  Stampian.  The  lower 
of  the  English  samples  (H23)  is  dominated  by  Adnatosphaeridium  reticulense,  which  is  also  a 
dominant  form  in  the  Argile  Verte.  The  upper  English  sample  (H24)  is  dominated  by  Phthanoperi- 
dinium  fiebile,  which  is  restricted  to  this  horizon  in  England  and  has  also  been  found  to  be 
abundant  in  the  Argile  Verte  by  Chateauneuf  (1978)  and  recorded  by  him  under  the  name  of 
Geiselodinium  cf.  geiseltalense.  This  strongly  suggests  a correlation  between  the  Nematura  Band  and  a 
horizon  within  the  Argile  Verte  de  Romainville. 

The  Upper  Hamstead  Beds  can  be  correlated  with  the  Calcaire  de  Sannois  and  the  lower  part  of  the 
Marnes  a Huitres.  This  correlation  is  supported  by  the  appearance  of  Gerdiocysta  conopeum 
( = Cyclonephelium  reticulosum  Gerlach,  Chateauneuf  1978),  W.  gochtii  (Chateauneuf,  pers.  comm.), 
P.  amoenum,  and  the  increased  abundances  of  W.  symmetrica  and  Pentadinium  taenigerum 
(Chateauneuf  1978)  in  both  areas. 

The  overlying  Sables  de  Fontainebleau  have  a rich  and  varied  dinoflagellate  assemblage  with 
species  such  as  Chiropteridium  lobospinosum  and  C.  partispinatum  (Chateauneuf  1978).  In  England 
there  is  no  representative  of  this  assemblage,  which  has  marked  similarities  to  those  from  the 
Rupelian  of  Germany  (Benedek  1972). 

Belgium 

Weyns  (1970)  described  two  assemblages  from  the  Sables  de  Grimmertingen  (Lower  Tongrian).  He 
listed  forty-seven  forms  of  dinoflagellate  cysts.  Of  these  thirty-six  are  apparently  present  in  the 
Middle  Headon  Beds,  and  the  assemblages  have  a general  similarity,  particularly  in  the  prominence 
of  Homotryblium  and  Spiniferites. 

In  comparison  with  the  Hamstead  Beds  assemblages,  there  are  major  differences.  The  many 
species  appearing  for  the  first  time  in  the  Hamstead  Beds  are  not  listed  in  Weyns’s  assemblages.  Only 
a few  of  the  species  listed  by  Weyns  appear  to  have  stratigraphic  significance.  Glaphyrocysta  micro- 
fenestrata  (=  C.  semicirculatum  in  Weyns)  does  not  appear  until  late  in  the  Chama  Beds  of  the 


LIENGJARERN  ET  AL.\  EOCENE/OLIGOCENE  DINOFL AGELL ATES 


497 


Bartonian  (Bujak  1976).  G.  exuberans  ellipsoidalis  and  Areosphaeridium  diktyoplokus  are  absent 
above  the  Middle  Headon  Beds.  The  correlation  that  best  fits  these  circumstances  is  between  the 
Sables  de  Grimmertingen  and  the  Middle  Headon  Beds.  This  is  in  agreement  with  recent  work  on  the 
nanoplankton  correlation  (Cavelier  1975).  A notable  difference  between  the  Belgian  and  English 
assemblages  is  the  presence  of  Leptodinium  and  Nematosphaeropsis  in  the  former.  These  are  forms 
found  to  be  more  prominent  in  open-sea  situations. 

Two  samples,  one  from  20  m and  the  other  from  30  m above  the  base  of  the  Rupel  Clay  in  the 
type  section,  yielded  rich  dinoflagellate  assemblages.  These  showed  marked  similarities  to  those  from 
the  Upper  Hamstead  Beds,  in  particular  containing  W.  gochtii.  However,  they  also  contain  C.  lobo- 
spinosum,  C.  partispinatum  and  other  species  which  are  not  present  in  the  Isle  of  Wight,  but  are 
characteristic  of  the  Sables  de  Fontainebleau  in  the  Paris  Basin,  and  the  Rupelton  in  Germany. 
These  samples  are  clearly  younger  than  any  from  the  Isle  of  Wight. 


THE  EOCENE/OLIGOCENE  BOUNDARY  IN  THE  ISLE  OF  WIGHT 

Establishment  of  a standard  for  this  stratigraphic  boundary  is  the  subject  of  continuing  debate.  In 
France,  it  has  commonly  been  placed  at  the  base  of  the  Stampian  Stage,  i.e.  at  the  base  of  the 
Argile  Verte  de  Romainville  (Chateauneuf  1978).  Accepting  this,  the  correlations  between  the  Isle  of 
Wight  succession  and  the  Paris  Basin  based  on  dinoflagellates  indicate  that  the  boundary  lies  closely 
below  the  Nematura  Band.  The  boundary  clearly  lies  between  the  Nematura  Band  and  the  Middle 
Headon  Beds.  The  Oyster  Bed,  although  it  has  a poor  assemblage,  has  greater  similarity  to  the 
Headon  Beds  than  to  the  succeeding  assemblage. 

Therefore,  if  the  French  view  is  accepted  the  boundary  lies  between  the  base  of  the  Nematura 
Band  and  the  top  of  the  Oyster  Bed.  Since  the  Argile  Verte  de  Romainville  marks  the  first 
important  marine  incursion  after  the  episode  of  the  Marnes  a Lucines  it  seems  very  likely  that  the 
Nematura  Band  represents  the  same  transgression.  The  Bembridge  Marls  then  correlate  with  the 
Supra-  and  Upper  Gypsiferous  Groups  (1st  and  2nd  mass)  and  the  Osborne  Beds  with  the  3rd  mass 
of  gypsum.  The  base  of  the  Oligocene  could  conveniently  be  taken  at  the  base  of  the  Hamstead  Beds, 
some  9 km  below  the  Nematura  Band. 

An  alternative,  widely  held,  view  is  that  the  base  of  the  Oligocene  originally  selected  in  Germany 
should  be  adopted.  This  is  marked  by  the  transgression  associated  with  the  Latdorf  (Lattorf)  Sands 
(NP21),  which  correlate  readily  with  the  Sables  de  Grimmertingen  in  Belgium. 

Dinoflagellates  have  not  been  described  from  the  Latdorf  Sands,  but  from  the  Sable  de  Grimmer- 
tingen assemblages  very  like  those  from  the  Middle  Headon  Beds  have  been  described  by  Weyns 
(1970).  If  this  correlation  is  accepted  the  Middle  Headon  Beds  would  be  Oligocene.  However,  the 
Brockenhurst  Bed  has  given  evidence  of  an  NP20  age,  which  indicates  that  the  base  should  be 
higher.  There  is,  however,  no  apparent  break  in  the  Middle  Headon  Beds  sequence,  only  a progressive 
increase  in  terrigenous  influence  in  the  Venus  Beds  (samples  WC22-25).  No  suitable  location  for  a 
boundary  is  evident. 

The  next  marine  incursion  in  the  Isle  of  Wight  succession,  the  Bembridge  Oyster  Bed,  did  not 
yield  any  dinoflagellates  of  much  value  in  correlation.  Those  that  are  present  are  not  inconsistent  with 
a correlation  with  the  Sables  de  Grimmertingen  and  consequently  with  the  placing  of  the  base  of  the 
Oligocene  immediately  above  the  Bembridge  Limestone,  as  is  done  by  Curry  et  al.  (1978). 


Acknowledgements.  We  particularly  thank  Dr.  J.  Bujak  for  information  regarding  the  Barton  Beds  and  for 
assisting  Dr.  Liengjarern  in  the  field;  Dr.  J.  J.  Chateauneuf  for  much  unpublished  data  on  the  Paris  Basin;  and 
Professor  D.  Curry  for  helpful  comments.  Dr.  Liengjarern  acknowledges  the  support  of  a Columbo  Plan 
Scholarship  enabling  her  to  do  this  research.  The  collections  are  housed  in  the  Department  of  Geology, 
University  of  Sheffield. 


PALAEONTOLOGY,  VOLUME  23 


REFERENCES 

benedek,  p.  N.  1972.  Phytoplanktonten  aus  dem  Mittel-und  Oberoligozan  von  Tonisberg  (Niederrheingebiet). 
Palaeontographica,  Abt.  B.,  137,  1-71. 

brideaux,  w.  w.  and  mcintyre,  d.  j.  1973.  Lunatadinium  dissolution  gen.  et  sp.  nov.,  a dinoflagellate  cyst  from 
Lower  Cretaceous  rocks,  Yukon  Territory  and  northern  District  of  Mackenzie.  Bull.  Can.  Petrol.  Geol.  21, 
395-402. 

bujak,  j.  p.  1973.  Microplankton  from  the  Barton  Beds  of  the  Hampshire  Basin,  England.  Ph.D.  thesis 
(unpubl.),  455  pp.,  University  of  Sheffield. 

— 1976.  An  evolutionary  series  of  late  Eocene  dinoflagellate  cysts  from  southern  England.  Mar.  Micro- 
paleontology, 1,  101-117. 

Cavelier,  c.  1975.  Le  diacronisme  de  la  zone  a Ericsonia  subdisticha  (Nannoplankton)  et  la  position 
de  la  limite  Eocene-Oligocene  en  Europe  et  en  Amerique  du  Nord.  Bull.  B.R.G.M.  (2),  4,  3,  201  - 
225. 

chateauneuf,  j-j.  1978.  Upper  Eocene  and  Oligocene  Dinophyceae  of  the  Paris  Basin  (France).  Proc.  4th 
Internat.  Palynological  Conf.,  Lucknow  (in  press). 

costa,  l.  i.  and  downie,  c.  1976.  The  distribution  of  the  dinoflagellate  Wetzeliella  in  the  Palaeogene  of  north- 
western Europe.  Palaeontology,  19,  591-614. 

CURRY,  D.,  ADAMS,  C.  G.,  BOULDER,  M.  C.,  DILLEY,  F.  C.,  EAMES,  F.  E.,  FUNNELL,  B.  M.  and  WELLS,  M.  K. 

1978.  A correlation  of  Tertiary  rocks  in  the  British  Isles.  Geol.  Soc.  Lond.  Special  Report,  No.  12, 
72  pp. 

daley,  B.  1973.  The  palaeoenvironment  of  the  Bembridge  Marls  (Oligocene)  of  the  Isle  of  Wight,  Hampshire. 
Proc.  Geol.  Ass.  84,  83-93. 

drugg,  w.  s.  1970.  Some  new  genera,  species,  and  combinations  of  phytoplankton  from  the  Lower 
Tertiary  of  the  Gulf  Coast  USA.  Proc.  North  Amer.  Paleont.  Convention,  Chicago,  1969,  G,  809- 
843. 

eisenack,  a.  1964.  Katalog  der  fossilen  Dinoflagellaten,  Hystrichospharen  und  verwandten  Mikrofossilien. 

Band  1.  Dinoflagellaten,  E.  888  pp.,  Schweizerbart’sche  Verlagsbuchhandlung,  Stuttgart. 

EVITT,  w.  r.,  LENTIN,  j.  k.,  millioud,  M.  E.,  stover,  L.  E.  and  williams,  G.  l.  1977.  Dinoflagellate  cyst 
terminology.  Geol.  Surv.  Pap.  Can.  76-24,  1-11. 

gerlach,  E.  1961.  Mikrofossilien  aus  dem  Oligozan  und  Miozan  Nordwestdeutschlands,  unter  besonderer 
Beriicksichtigung  der  Hystrichosphaeren  und  Dinoflagellaten.  Neues  Jb.  Geol.  Pal'aontol.  Abh.  5,  112, 
143-228. 

Harris,  w.  k.  1973.  Tertiary  non-marine  dinoflagellate  cyst  assemblages  from  Australia.  Spec.  Pubis,  geol. 
Soc.  Aust.  4,  159-166. 

haskins,  c.  w.  1969.  Tertiary  Ostracoda  from  the  Isle  of  Wight  and  Barton,  Hampshire,  England.  Part  IV. 
Revue  Micropaleont.  12,  149-170. 

keen,  M.  c.  1972.  The  Sannoisian  and  some  other  upper  Palaeogene  Ostracoda  from  north-west  Europe. 
Palaeontology,  15,  267-325. 

— 1977.  Ostracod  assemblages  and  the  depositional  environments  of  the  Headon,  Osborne  and  Bembridge 
Beds  (Upper  Eocene)  of  the  Hampshire  Basin.  Ibid.  20,  405-445. 
lentin,  J.  K.  and  williams,  G.  L.  1975.  A monograph  of  fossil  peridinioid  dinoflagellate  cysts.  Bedford  Institute 
Oceanography,  Report  Bl-R-75-16,  1-237. 

liengjarern,  m.  1973.  Dinoflagellate  cysts  and  acritarchs  from  the  Oligocene  Beds  of  the  Isle  of  Wight.  Ph.D. 
thesis  (unpubl.),  220  pp.,  University  of  Sheffield. 

mcintyre,  d.  J.  1975.  Morphologic  changes  in  Deflandrea  from  a Campanian  section,  District  of  Mackenzie, 
N.W.T.,  Canada.  Geosci.  Man,  11,  61-76. 

morgenroth,  p.  1966a.  Mikrofossilien  und  Konkretionen  des  nordwesteuropaischen  Untereozans.  Palaeonto- 
graphica, Abt.  B.,  119,  1-53. 

— 1966 b.  Neue  in  organischer  Substanz  erhaetene  Mikrofossilien  des  Oligozans.  Neues  Jb.  Geol.  Palaont. 
Abh.  127,  1-12. 

Murray,  j.  w.  and  wright,  c.  a.  1974.  Palaeogene  Foraminiferida  and  palaeoecology,  Hampshire  and  Paris 
Basins  and  the  English  Channel.  Spec.  Pap.  Palaeontology,  14,  1-171. 
stover,  l.  E.  1977.  Oligocene  and  early  Miocene  dinoflagellates  from  Atlantic  Corehole  5/5B,  Blake  Plateau. 
Am.  Assoc.  Stratigr.  Palynol.,  Contrib.  Ser.  5A,  66-89. 

— and  evitt,  w.  R.  1978.  Analyses  of  Pre-Pleistocene  organic  walled  Dinoflagellates.  Stanf.  Univ.  Pubis, 
Geol.  Sciences,  15,  1-300. 


LIENGJ ARERN  ET  AL..  EOCENE/OLIGOCENE  DINOFLAGELLATES 


499 


weyns,  w.  1970.  Dinophycees  et  acritarches  des  ‘Sables  de  Grimmertingen’  dans  leur  localite-type,  et  les 
problemes  stratigraphiques  du  Tongrien.  Bull.  Soc.  beige  Geol.  Paleont.  Hydrol.  79,  247-268. 
williams,  G.  l.,  sarjeant,  w.  a.  s.  and  kidson,  e.  j.  1973.  A glossary  of  the  terminology  applied  to  dino- 
flagellate  amphiesmae  and  cysts  and  acritarchs.  Am.  Assoc.  Stratigr.  Palynol.,  Contrib.  Ser.  2,  1-222. 

M.  LIENGJ  ARERN 
L.  COSTA 
C.  DOWNIE 

Department  of  Geology 

Manuscript  received  21  December  1978  University  of  Sheffield 

Revised  manuscript  received  18  July  1979  Sheffield  SI  3JD 


DICTYODORA  FROM  THE  SILURIAN  OF 
PEEBLESSHIRE,  SCOTLAND 

by  m.  j.  benton  and  n.  h.  trewin 


Abstract.  The  meandering  trace  fossil  Dictyodora  Weiss,  1 884  occurs  in  deep  water  greywacke/shale  sequences 
in  the  Gala  Group  (lower  Silurian)  of  Thornylee  and  Grieston  Quarries,  Galashiels.  Two  species  are  recognized; 
D.  scotica  (M‘Coy,  1851)  and  D.  tenuis  (M‘Coy,  1851);  the  former  is  distinguished  by  a more  regular  meandering 
form.  These  traces  were  originally  named  Crossopodia  scotica  and  Myrianites  tenuis.  It  is  suggested  that 
C.  scotica  be  rejected  as  the  type  species  of  Crossopodia. 


Thornylee  Quarry  (Grid  ref.  NT  4200  3635)  (formerly  spelt  Thornyly,  Thorney  Lee, 
Thornielee,  Thornilee)  is  situated  on  the  north  bank  of  the  River  Tweed,  8 km  east  of  Galashiels  and 
8 km  west  of  Innerleithen.  The  quarry  is  located  on  a steep  slope  above  a layby  on  the  A72 
(Peebles-Galashiels)  road.  Between  the  quarry  and  the  road  is  a dismantled  railway  with  cuttings 
which  provide  a 300  m long  section  through  Upper  Llandovery  greywackes  and  shales  (Gala  Group 
of  Lapworth  1870).  The  first  geological  description  of  Thornylee  was  given  by  Nicol  (1850)  who 
noted  some  graptolites  and  abundant  ‘annelid  impressions’. 

Grieston  Quarry  (NT  3130  3618)  was  also  described  by  Nicol  (1850),  who  noted  the  abundant 
graptolite  fauna  and  the  trace  fossils.  More  recently  the  fauna  and  sediments  of  this  quarry  have  been 
described  by  Toghill  and  Strachan  (1970)  and  Trewin  (1979).  The  thin  greywackes  and  shales  of 
Grieston  also  lie  within  the  top  of  the  Gala  Group  of  Lapworth  (1870),  but  are  not  exactly  the  same 
age  as  those  at  Thornylee  on  the  basis  of  the  graptolite  fauna. 

This  study  stemmed  from  work  on  H.  A.  Nicholson’s  trace  fossil  collection  in  Aberdeen  (Benton 
and  Trewin  1978).  The  following  descriptions  are  based  on  large  collections  made  at  Thornylee  and 
Grieston  in  April  and  June,  1977.  Comparisons  have  been  made  with  the  type  material  of  M‘Coy  and 
Nicholson.  Repository  abbreviations  used  are:  AUGD,  Aberdeen  University,  Department  of 
Geology  and  Mineralogy  Palaeontology  Collection;  BMNH,  British  Museum  (Natural  History); 
GSM,  Geological  Survey  Museum,  I.G.S.,  London;  HM,  Hunterian  Museum,  Glasgow;  SM, 
Sedgwick  Museum,  Cambridge. 


DEPOSITIONAL  ENVIRONMENT  AND  ASSOCIATED  FAUNA 

At  both  localities  deep  water,  interbedded  greywacke/shale  sequences  are  exposed  in  which  the 
coarser  lithologies  are  of  turbidite  origin.  The  trace  fossils  at  Thornylee  are  more  abundant  in  the 
shale-rich  parts  of  the  sequence  rather  than  in  association  with  greywacke  beds.  There  seems  to  be  a 
greater  frequency  of  meandering  traces  in  the  purple  rather  than  the  green  shales.  At  Grieston  the 
greywackes  are  fine-grained  and  contain  abundant  ripple-lamination,  possibly  the  results  of 
reworking;  other  beds  are  characterized  by  numerous  transported  graptolites  which  produced 
delicate  tool  marks  on  bed  bases  (Trewin  1979).  The  greywackes  at  Thornylee  are  usually  medium 
grained,  graded,  and  sometimes  show  tool  marks  and  load  casts  on  the  sharp  bed  bases.  Internally, 
Bouma  sequences  of  structures  are  frequently  seen.  The  general  aspect  of  the  lithofacies  is  of  a low- 
energy  turbidite  environment  with  thin  greywacke  turbidites  and  abundant  shale. 

At  both  localities  graptolites  are  present  but  they  are  much  more  abundant  in  the  finer  grained 
rocks  of  Grieston  Quarry,  where  the  majority  have  been  transported  and  deposited  in  thin  turbidites. 


[Palaeontology,  Vol.  23,  Part  3, 1980,  pp.  501-513.1 


502 


PALAEONTOLOGY,  VOLUME  23 


Tail  spines  of  Ceratiocaris  occur  at  Grieston,  but  no  other  fauna  was  noted.  The  ichnofauna 
dominated  by  meandering  feeding  burrows  is  typical  of  deep  water  muds  and  belongs  to  Seilacher’s 
Nereites  facies. 


THE  ICHNOFAUNA 

Introduction.  The  ichnofauna  is  dominated  by  the  meandering  burrows  of  two  species  of  Dictyodora, 
which  are  described  below.  The  small  burrow  Caridolites  Etheridge,  Woodward  and  Jones,  1890  is 
common  at  both  localities.  Rare  examples  of  Nereites  were  found  at  Thornylee  and  stuffed  burrows, 
cf.  Planolites,  are  also  present.  The  meandering  traces  are  described  below  with  more  emphasis  placed 
on  Dictyodora  scotica  in  view  of  its  taxonomic  importance.  A redescription  is  given  of  Caridolites  and 
the  association  with  Nereites  briefly  discussed. 

Genus  dictyodora  Weiss,  1884 
Taxonomic  discussion  of  Dictyodora 

Geinitz  (1867)  founded  the  species  Dictyodora  liebeanum  for  a ‘plant’  from  the  Culm  (Lower 
Carboniferous)  of  Gera,  East  Germany,  and  Weiss  (1884a,  b)  proposed  the  genus  Dictyodora  for  this 
species.  He  was  unable  to  decide  if  it  was  of  plant  or  animal  origin. 

Zimmermann  (1889,  1891)  discussed  the  taxonomic  problems  associated  with  German  Carbon- 
iferous Dictyodora,  noticing  that  as  with  the  British  examples,  different  horizontal  (bedding  parallel) 
sections  had  been  given  distinct  names  at  different  times.  Zimmermann  (1892)  gave  a detailed  account 
of  the  type  species  D.  liebeana,  and  considered  that  the  vertical  wall  contained  no  infill,  but  noted 
longitudinal  and  oblique  streaks.  Zimmermann  noted  that  the  wall  tends  to  slope  inwards  towards 
the  top,  giving  tighter  loops  than  those  of  the  basal  burrow,  but  was  puzzled  by  walls  intersecting 
without  disturbance.  D.  liebeana  has  vertical  walls  up  to  1 80  mm  high  and  a well-defined  over-all  cone 
shape  distinguishing  it  from  D.  scotica  and  D.  tenuis.  Zimmermann  (1892)  briefly  described  a species, 
D.  hercynica,  which  has  a looser  structure  and  walls  1 -3  cm  high,  found  in  the  Upper  Devonian  of  the 
Harz  mountains.  It  has  apparently  not  been  figured. 

D.  simplex  Seilacher,  1955  from  the  Lower  Cambrian  of  the  Salt  Range  of  Pakistan  is  a simple, 
loose  structure  about  6 mm  deep.  However,  this  is  a structure  built  from  successive  sloping  layers  and 
Seilacher  proposed  that  the  trace  was  produced  by  a worm-like  animal  travelling  through  the 
sediment  in  an  oblique  position.  There  is  no  basal  burrow  in  Seilacher’s  reconstruction  and  the 
‘vertical  wall’  is  of  equal  width  from  top  to  bottom.  We  consider  that  these  differences  are  sufficient  to 
exclude  D.  simplex  from  the  genus  Dictyodora.  No  alternative  generic  assignment  is  suggested 
without  examination  of  the  original  material. 

Seilacher  (1967,  p.  77)  figured  a Dictyodora  evolutionary  sequence  from  relatively  loosely 
structured  forms  in  the  Lower  Palaeozoic  to  tightly  spiralling  patterns  in  the  Carboniferous.  In  grade 
of  organization,  D.  tenuis  appears  similar  to  Seilacher’s  most  primitive  type  (a)  and  D.  scotica  is 
slightly  more  advanced. 

Pfeiffer  (1959)  reviewed  previous  work  on  D.  liebeana  and  gave  good  three-dimensional 
reconstructions  of  Carboniferous  examples.  Muller  (1962)  also  described  the  morphology  of  German 
Lower  Carboniferous  Dictyodora  in  detail  with  many  figures,  and  Ruchholz  (1967)  gave  further 
examples  from  the  Harz  mountains.  Pfeiffer  (1968)  gave  a synonomy  list  for  D.  liebeana  (Geinitz, 
1867).  Muller  (1971)  discussed  the  formation  of  Dictyodora  meanders,  emphasizing  that  the  trace  was 
a feeding  structure  formed  relatively  rapidly,  since  the  basal  burrow  does  not  change  in  diameter  in 
any  single  specimen  and  since  it  maintains  a constant  depth  and  does  not  rise  gradually  to  keep  up 
with  sedimentation. 

There  is  thus  an  extensive,  mainly  German,  literature  on  Dictyodora  which  establishes  the 
characteristic  features  of  the  genus  as  the  meandering  basal  burrow  and  the  dorsal  striated  wall.  The 
species  D.  scotica,  described  below,  has  previously  been  given  the  name  Myrianites  tenuis  for  sections 
for  the  vertical  wall  and  Crossopodia  scotica  for  the  basal  burrow. 


BENTON  AND  TREWIN:  DICTYODORA 


503 


The  genus  Myrianites  MacLeay,  1839  was  established  for  a meandering  track  with  small  leaf-like 
extensions  at  the  sides.  The  type  species,  M.  macleayii  Murchison,  1839  (type  specimen:  GSM  Geol. 
Soc.  Coll.  6824)  appears  to  be  a small  Nereites.  Species  from  Spain  described  by  Delgado  (1910)  as 
Myrianites  are  certainly  Dictyodora  but  are  not  described  or  figured  well  enough  to  establish 
synonomy  with  the  material  described  here. 

M‘Coy  (1851a,  b ) founded  the  species  M.  tenuis  based  on  specimens  of  small  meandering  traces 
from  Grieston  Quarry.  Nicholson  (1978,  pp.  42,  43)  identified  wall  sections  of  D.  scotica  from 
Thornylee  as  M.  tenuis,  but  the  specific  name  tenuis  is  retained  here  for  M‘Coy’s  original  material 
redescribed  below  as  D.  tenuis. 

M‘Coy  (1851a,  b)  also  founded  the  genus  Crossopodia  for  two  Silurian  trace  fossils.  C.  lata  from 
Llandeilo,  Wales,  is  a 2 cm  wide  trail  with  clear  transverse  striations  and  a ‘fringe’  which  better 
resembles  the  Crossopodia  of  modern  usage.  C.  scotica,  however,  is  the  form  redescribed  here  as 
D.  scotica  and  M‘Coy’s  type  (SM  A45575a-c)  clearly  shows  the  diagnostic  features  (text-fig.  2).  The 
figure  of  the  type  of  C.  scotica  in  M‘Coy  18516,  pi.  ID,  fig.  15,  appears  to  be  a composite  of  the  three 
specimens  SM  A45575a-c.  Fortunately  all  are  of  the  same  species  and  A45575a  is  more  suitable  as  the 
lectotype  showing  well  all  the  major  features.  M‘Coy’s  figure  has  been  reversed  in  the  engraving 
process.  Unfortunately,  Hantzschel  (1962,  p.  W189)  designated  C.  scotica  as  the  type  species  of 
Crossopodia  and  repeated  this  with  a mislabelled  figure  of ‘C.  scotia'  (sic)  in  Hantzschel  (1975,  fig.  34, 
2b).  This  figure  is  derived  from  Schimper  and  Schenk  (1879,  p.  52,  fig.  40)  and  is  clearly  not  the 
C.  scotica  of  M‘Coy  (1851a,  b)  and  Nicholson  (1978). 

In  order  to  preserve  the  normally  accepted  usages  of  Crossopodia  and  Dictyodora  we  propose  that 
C.  scotica  be  rejected  as  the  type  species  of  Crossopodia.  C.  lata  M‘Coy  (1851)  (type  specimen 
SM  A37733)  would  then  become  the  type  species  of  Crossopodia.  An  application  to  this  effect  will  be 
made  to  the  I.C.Z.N.  or  other  appropriate  body,  when  agreement  has  been  achieved  on  the  rules  of 
trace  fossil  nomenclature.  Further  revision  of  the  genus  Crossopodia  is  required,  but  is  outside  the 
scope  of  this  paper. 


Dictyodora  scotica  (M‘Coy,  1851) 

Text-figs.  1,  2,  3 

v*  1851a  Crossopodia  scotica  M‘Coy,  p.  395. 

v*  18516  Crossopodia  scotica  M‘Coy;  M‘Coy,  p.  130,  pi.  ID,  fig.  15. 

71855  Crossopodia  scotica  M‘Coy;  Harkness,  p.  475. 
non  1879  Crossopodia  scotica  (M‘Coy);  Schimper  and  Schenk,  p.  52,  fig.  40. 
non  1962  Crossopodia  scotia  (M‘Coy)  (sic);  Hantzschel,  p.  W189,  fig.  118,  2. 
non  1975  Crossopodia  scotia  (M‘Coy)  (sic);  Hantzschel,  p.  W54,  fig.  34,  2b. 
vl978  Crossopodia  scotica  M‘Coy;  Nicholson,  p.  36,  pi.  3,  fig.  1,  pi.  6. 

vl978  Myrianites  tenuis  M‘Coy;  Nicholson,  p.  42,  text-fig.  7,  non  pi.  4,  fig.  1.  [The  same  specimen  as  in 
Benton  and  Trewin  1978,  pi.  2,  fig.  2.] 
vl978  Crossopodia  scotica  M‘Coy;  Benton  and  Trewin,  p.  8,  pi.  2,  fig.  1. 

Lectotype.  Here  designated,  SM  A45575a,  the  original  of  M‘Coy  (18516,  pi.  ID,  fig.  15).  Gala  Group,  Upper 
Llandovery,  lower  Silurian,  Thornylee  Quarry,  nr.  Innerleithen,  Peeblesshire,  Scotland.  Refigured  here, 
text-fig.  2. 

Other  material.  More  than  two  hundred  examples  from  the  type  locality,  a representative  selection  of  which  are 
catalogued  as  AUGD  10693  to  10710.  Also:  AUGD  8819,  8820,  10606,  10723,  Mus.  Coll.  956,  957;  BMNH 
39451,  58169  (1,  2);  GSM  104247,  104249,  104250,  RU  2970;  HM  X871/1-2,  X1003/1-7. 

Description.  The  burrow  system  illustrated  in  text-fig.  1 consists  of  a basal  burrow,  generally  preserved  with  a 
lenticular  cross  section,  and  having  a vertical  or  inclined  longitudinal  wall  arising  from  the  dorsal  mid-line  of  the 
basal  burrow.  The  basal  burrow  varies  from  1-5-6  mm  wide  and  up  to  3 mm  high  in  slate  lithologies,  but  when 
developed  in  fine  sand  may  have  a nearly  circular  cross  section  due  to  the  small  degree  of  compaction.  The  wall  is 
up  to  13  mm  high  and  tapers  upwards  from  a width  of  1 -2  mm  at  the  base.  The  taper  is  most  rapid  in  small 
examples.  The  typical  burrow  system  (text-fig.  3c,  d,  e)  consists  of  5-10  parallel  meanders  each  10-80  mm  long 


504 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  1.  Scale  bars  10  mm  at  front  faces  of  figures.  Arrows  indicate  direction  of  travel  of  Dictyodora 
organism,  a,  general  morphology  of  Dictyodora  meanders  showing  basal  burrow  and  wall;  wall  curves  inwards 
at  meander  bends.  B,  section  of  burrow  to  show  features  of  burrow  and  wall  fill,  horizontal  striations  and  curved 
vertical/oblique  striations  of  wall  surface,  c,  block  diagram  illustrating  different  preservational  aspects  of  the 
burrow  in  plan  and  section;  a,  narrow  sections  at  top  of  wall;  b,  wider  sections  near  base  of  wall;  c,  convex  top  of 
basal  burrow  with  base  of  wall  fill  preserved  on  top;  d,  concave  impression  of  underside  of  burrow  with  fill 
removed,  a weak  median  ridge  may  be  present;  e,  smaller  example  showing  effect  of  sectioning  the  inclined  wall 
at  meander  turn;  /,  juvenile  burrow  in  section.  The  style  of  ripples  and  fine  parallel  lamination  present  is  also 
illustrated  on  the  front  face  of  c. 


BENTON  AND  TREWIN:  DICTYODORA 


505 


(usually  30-50  mm)  and  internally  measured  at  basal  burrow  level  as  0-20  mm  apart  (usually  5-15  mm).  Where 
successive  meanders  touch,  a tight  turning  circle  is  present  at  the  meander  turn.  The  meanders  may  also  be 
irregular  and  broad  as  in  text-fig.  3a,  b.  The  relevant  features  of  the  type  specimen  are  illustrated  in  text-fig.  2. 

The  burrow  shows  various  preservational  aspects  (text-fig.  lc)  dependent  on  the  level  at  which  it  is  sectioned. 
Sections  of  the  wall  appear  as  meandering  lines  up  to  2 mm  wide,  occasional  sharp  turns  are  seen  in  sections  close 
to  the  top  of  the  wall  (text-fig.  3e)  but  nearer  the  basal  burrow  the  wall  displays  smooth  curves.  The  wall  has  a 
finite  thickness  and  the  burrow  may  break  either  side  of  the  wall  as  shown  in  text-fig.  3b.  Sections  at  the  top  of  the 
basal  burrow  show  the  entire  infill  with  a median  ridge  marking  the  base  of  the  wall  (text-fig.  lc).  Specimens 
showing  the  lower  surface  of  the  basal  burrow  display  a smooth  groove  which  is  sometimes  double,  with  a weak 
median  ridge  (text-fig.  lc).  The  burrow  may  also  split  within  the  burrow  fill  giving  very  little  relief  to  the 
preserved  trace.  Internally,  a distinct  pattern  is  frequently  seen  in  polished  or  etched  cross-sections  of  the  burrow 
fill  resulting  from  reorientation  of  platey  minerals  (text-fig.  1b). 


text-fig.  2.  Sketch  of  lectotype  of  Dictyodora  scotica,  SM  A45575a  showing  the  lower  surface  of  the  specimen. 
Trace  A shows  the  typical  meander  pattern.  Most  of  the  specimen  displays  the  lower  surface  of  the  burrow  but  at 
a the  burrow  fill  is  broken  out  to  show  a mould  of  the  upper  surface  of  the  basal  burrow.  The  wall  of  A is  5 mm 
high  and  is  not  seen  on  the  top  of  the  slab.  Trace  B is  larger  than  A and  later  since  it  clearly  crosses  A.  At  b the 
transition  from  basal  burrow  to  wall  can  be  seen.  The  wall  passes  through  the  full  8 mm  thickness  of  the  slab  and 
is  seen  on  the  top  of  the  specimen  (not  illustrated). 

The  burrows  are  indistinct  in  places  due  to  the  presence  of  several  crossing  burrows,  and  fracture  irregularities 
on  the  surface  of  the  slab  which  have  been  omitted  for  clarity. 


The  wall  is  normally  vertical  above  straight  stretches  of  burrow,  but  curves  inwards  at  meander  bends  (text- 
figs.  1a,  c,  3b,  e).  Fine  bedding  parallel  striations  are  present  on  the  surface  of  the  wall  closely  spaced  at  4 per  mm. 
A similar  bedding  parallel  banding  due  to  platey  mineral  orientation  occurs  within  the  wall  fill,  and  is  not  related 
to  sedimentary  laminae.  Curved  vertical/oblique  striations  are  also  present  on  the  wall  surface  normally  spaced 
at  3-5  per  mm.  Internally  the  wall  may  show  fine  curved  structures  marked  by  reoriented  platey  minerals  and 
resembling  backfill  within  the  wall  (text-figs.  1 b,  3b).  Detailed  observation  of  features  is  difficult  in  the  wall  fill  but 
it  is  likely  that  the  possible  backfill  structures  seen  normal  to  bedding  occur  between  the  bedding  parallel  bands. 

The  smallest  forms  recognized  have  a basal  burrow  1.5  mm  wide  and  a wall  only  1 mm  high,  and  a full 
gradation  exists  up  to  the  larger  forms  with  a progressive  increase  in  wall  height  relative  to  burrow  width  (text- 
fig.  4).  Detailed  measurement  of  the  morphology  and  meander  patterns  of  over  170  specimens  using  principal 
components  analyses  failed  to  differentiate  any  groups  with  significantly  different  characters,  and  we  consider 
that  all  the  meandering  burrows  of  this  type  are  growth  stages  of  a single  species. 


text-fig.  3.  Dictyodora  scoticcr,  examples  of  burrow  morphology,  a,  irregular  meanders  (section  of  burrow  wall) 
with  example  of  avoidance  of  previously  formed  burrow  at  a,  AUGD  10693.  b,  irregular  burrow  which  crosses 
previously  formed  burrow;  plan  view  shows  wall  above  basal  burrow  to  be  partly  broken  away,  and  inward  slope 
of  wall  at  meander  curves;  thickness  of  slab  10  mm;  AUGD  10697.  c,  D,  typical  regular  meander  forms,  hooked 
ends  to  meanders  seen  in  c;  both  on  AUGD  10694.  e,  plan  view  of  basal  burrow  (stipple)  and  position  of  top  of 
wall  (solid  line);  sharp  bends  present  at  top  of  wall  become  smooth  curves  at  lower  levels  close  to  the  basal 
burrow;  AUGD  10698.  All  examples  from  Thornylee  Quarry. 


BENTON  AND  TREWIN:  DICTYODORA 


507 


Occurrence.  Dictyodora  scotica  is  common  at  Thornylee  Quarry  and  scarce  at  Grieston  Quarry.  It  is  probably 
common  in  the  Llandovery  strata  of  the  Southern  Uplands  since  Peach  and  Horne  ( 1 899)  mention  ‘ Crossopodia ’ 
and  ‘ Myrianites ’ from  at  least  twenty  localities  in  the  Galashiels-Hawick  region.  It  also  occurs  in  the  Llandovery 
of  Penwhapple  Glen,  Girvan  (Nicholson  and  Etheridge  1880,  pp.  304-318).  P.  Doughty  (pers.  comm.)  also 
records  Dictyodora  from  the  Silurian  of  Co.  Down,  Northern  Ireland. 


H 1 1 1 1 r 

1 2 3 4 5 6 W mm 


text-fig.  4.  Dictyodora  scotica.  Relationship  of  width  of  basal  burrow  W with  burrow 
height  H to  show  range  of  variation  and  the  relative  increase  in  wall  height  in  the  larger 
examples. 


Dictyodora  tenuis  (M‘Coy,  1851) 
Text-fig.  5 


v*  1851a  Myrianites  tenuis  M‘Coy,  p.  394. 
v*18516  Myrianites  tenuis  M‘Coy;  M‘Coy,  p.  130,  pi.  ID,  fig.  13. 
vl978  Myrianites  tenuis  M‘Coy;  Nicholson,  pi.  4,  fig.  1,  non  text-fig.  7. 
vl978  Myrianites  murchisoni  Emmons;  Nicholson,  p.  43,  pi.  5,  fig.  1 . 

Lectotype.  Here  designated,  SM  A45579a,  the  original  of  M‘Coy  (18516,  pi.  ID,  fig.  13).  Gala  Group,  Upper 
Llandovery,  Lower  Silurian,  Grieston  Quarry,  nr.  Innerleithen,  Peeblesshire,  Scotland  (text-fig.  5a). 

Other  material.  AUGD  9224, 10329,  10607, 10612,  and  10711  to  10720  from  Grieston  Quarry  and  AUGD  10710 
from  Thornylee  Quarry. 


508 


PALAEONTOLOGY,  VOLUME  23 


Description.  Dictyodora  with  broad  irregular  meanders,  as  in  text-fig.  5,  which  frequently  have  a secondary 
sinuosity  with  a wavelength  of  3-15  mm  which  may  develop  into  meanders  with  length  roughly  equal  to  breadth 
in  larger  examples.  The  basal  burrow  is  from  T5  to  3 mm  wide  and  the  wall  has  not  been  observed  to  exceed 
10  mm  in  height.  The  wall  is  0-2-0-7  mm  wide  and  striated  in  the  same  manner  as  in  D.  scotica.  Traces  range  from 
tiny  ‘scribbles’  (text-fig.  5e)  up  to  large  examples  as  in  text-fig.  5b,  d. 

Trace  endings  are  observed  as  in  text-fig.  5b  where  lengths  of  trace  as  short  as  10  mm  occur  between  inclined 
circular  burrows  3 mm  in  diameter;  other  traces  can  be  followed  for  over  200  mm  without  interruption. 

Discussion.  The  distinction  of  D.  tenuis  from  D.  scotica  can  be  made  on  maximum  size  and  on  the 
meandering  pattern,  which  is  more  regular  and  smooth  in  D.  scotica  compared  with  the  irregular 
meanders  with  secondary  sinuosity  displayed  by  D.  tenuis. 

In  the  past  specimens  displaying  sections  of  the  wall  have  been  identified  as  Myrianites  and 
specimens  showing  the  basal  burrow  as  Crossopodia  or  Nemertites.  The  specimens  from  Grieston 
called  M.  murchisoniby  Nicholson  (1978,  p.  43,  pi.  15,  fig.  1)  are  not  synonymous  with  the  American 
form  described  by  Emmons  (1844)  and  are  ascribed  here  to  D.  tenuis. 

Occurrence.  Common  in  the  Upper  Llandovery  ( griestonensis  Zone)  of  Grieston  Quarry,  nr.  Innerleithen, 
Peeblesshire,  and  also  present  in  association  with  much  commoner  D.  scotica  at  Thornylee  Quarry.  The  form 
illustrated  by  Raup  and  Seilacher  (1969,  fig.  la)  from  the  Ordovician  of  Barrancos,  Portugal,  appears  to  be 
D.  tenuis. 


THE  DICTYODORA  ANIMAL  AND  ITS  BEHAVIOUR 

The  meandering  burrow  of  Dictyodora  resembles  the  meandering  burrows  and  trails  produced  by 
worms  and  molluscs  efficiently  utilizing  an  area  as  a food  source.  The  tightly  packed  meanders  of 
Dictyodora  were  probably  formed  during  feeding,  and  the  looser  irregular  meanders  may  have  been 
the  result  of  searching  for  areas  rich  in  food.  We  assume  that  the  body  of  the  animal  occupied  the 
basal  burrow,  and  probably  progressed  by  peristaltic  movement.  Since  individual  burrows  cannot  be 
traced  from  small  to  large  size,  and  considering  that  the  burrows  are  sometimes  seen  to  end  by  rising 
through  the  sediment  it  is  likely  that  the  animal  moved  from  place  to  place  on  or  above  the  sediment 
surface.  Thus  the  burrows  are  considered  to  be  produced  by  short  periods  of  food  search  and 
utilization  at  a constant  level  within  the  sediment. 

The  animal  appears  to  have  maintained  contact  with  the  surface  by  means  of  an  organ  which  was 
responsible  for  the  production  of  the  striated  wall  on  the  dorsal  burrow  surface;  this  we  term  the  wall- 
organ  to  avoid  assumptions  implicit  in  the  use  of  known  zoological  terms  such  as  ‘siphon’.  The 
curved  vertical  striations  on  the  wall  and  the  fill  of  the  wall  indicate  that  the  wall-organ  moved 
regularly  through  the  sediments,  maintaining  a constant  convex-forward  edge  and  followed  the 
movement  of  the  animal  in  the  burrow;  thus  wall-organ  traces  occasionally  touch  or  cross  each  other 
while  the  corresponding  burrows  do  not. 

The  behaviour  of  animals  that  form  meandering  traces  has  been  discussed  by  several  authors. 
Seilacher  (1967)  suggested  that  the  Dictyodora  animal  measured  its  meander  length  by  the  length  of 
its  body.  It  maintained  contact  with  a previously  formed  burrow  (thigmotaxis)  until  its  body  was 
straight  and  then  the  animal  was  ‘programmed’  to  make  a sharp  U-turn  (homostrophy)  as  its  tail 
straightened,  and  to  follow  beside  the  last-formed  portion  of  the  burrow.  However,  this  explanation 
does  not  satisfactorily  explain  individual  burrows  where  meander  length  varies,  or  the  Carboniferous 
Dictyodora  where  the  meanders  spiral  out  from  a central  point,  each  meander  being  longer  than  its 
predecessor. 

Seilacher  based  his  interpretation  on  the  classic  work  of  Richter  (1924,  1928),  who  studied  the 
Cretaceous/Tertiary  Helminthoida  labyrinthica  Heer,  1865  which  forms  similar  meandering  feeding 
traces.  Richter’s  interpretation  differs  from  Seilacher’s  in  one  important  way:  he  defined  the 
homostrophic  turning  stimulus  as  caused  by  loss  of  contact  with  a former  trace  and  not  by  tail 
straightening.  The  animal  followed  a former  trace  and  could  at  times  curve  in  front  of  previous 
meander  ends  (e.g.  text-fig.  3c)  before  turning  back  when  it  lost  contact  with  disturbed  mud.  In  text- 
fig.  3 meander  length  varies  from  30  to  80  mm  and  was  clearly  not  measured  by  the  body  length  of  the 


BENTON  AND  TREWIN:  DICTYODORA 


509 


text-fig.  5.  Dictyodora  tenuis.  Examples  of  burrow  morphology  shown  by  sections  of  the  wall  of  the  burrow. 

a,  small  meandering  trace  with  irregular  meanders  showing  secondary  sinuosity;  part  of  lectotype  SM  45579a. 

b,  parts  of  typical  irregular  meanders,  together  with  short  lengths  of  burrow  terminated  by  inclined  sections  of 
basal  burrow;  AUGD  10719.  c,  d,  e,  irregular  meanders  of  various  sizes  to  show  variation  in  meander 

morphology;  c,  E AUGD  10716;  d AUGD  10718.  All  from  Grieston  Quarry. 


510 


PALAEONTOLOGY,  VOLUME  23 


animal.  The  reactions  of  the  animal  while  feeding  in  meanders  as  listed  by  Seilacher  (1967)  and  Raup 
and  Seilacher  (1969)  may  be  modified  to: 

(1)  Move  horizontally  keeping  within  a single  stratum  of  sediment  (? controlled  by  wall-organ  length); 

(2)  Always  keep  in  touch  with  previously  formed  burrow  while  feeding  (thigmotaxis); 

(3)  Never  come  closer  to  a previously  formed  burrow  than  a particular  distance  ‘d’  (phobotaxis); 

(4)  If  contact  is  lost  with  a former  burrow,  make  a 180°  turn  (homostrophy/strophotaxis). 

These  ‘rules’  appear  to  apply  reasonably  well,  and  obvious  cases  of  burrow  avoidance  can  be  found 
(text-fig.  3a).  Traces  made  by  individuals  at  different  levels  in  the  sediment  frequently  cross  each 
other,  but  the  basal  burrows  in  such  cases  are  normally  at  different  levels.  In  the  Thornylee  examples 
population  density  was  probably  low  and  thus  there  was  no  need  for  attempting  to  utilize  an  area 
more  than  once. 

If  the  meandering  burrows  are  formed  during  feeding  then  the  question  arises  of  how  feeding  was 
accomplished.  The  wall-organ  could  have  been  a food  collector  at  the  surface,  with  the  animal 
protected  in  its  burrow,  or  the  animal  could  have  fed  by  sediment  ingestion  at  burrow  level  leaving  the 
wall-organ  to  perform  a respiratory  function.  The  second  of  these  suggestions  seems  most  favourable 
since  the  basal  burrow  has  a definite  burrow  fill  which  corresponds  to  the  sediment  type  at  basal 
burrow  rather  than  surface  level.  The  apparently  passive  motion  of  the  wall-organ  does  not  accord 
with  a function  as  a feeding  organ,  and  it  is  more  likely  to  have  had  a respiratory  function  and  to  have 
controlled  burrow  depth. 

In  laminated  sediment  the  fill  of  the  wall  roughly  matches  the  characteristics  of  the  immediately 
adjacent  sediment,  with  only  slight  downward  movement  of  sediment  during  filling  occasionally  seen 
in  thin  section.  Thus  the  wall-organ  does  not  seem  to  have  had  a significant  sediment  transport 
function.  No  annulation  of  the  burrow  fill  is  seen  and  the  constant  fine  spacing  of  the  striations 
formed  by  the  wall-organ  would  seem  to  indicate  a slow  regular  movement  through  the  sediment.  The 
wall-organ  may  have  been  ciliated  to  facilitate  its  progress  through  the  sediment.  The  striations  and 
structured  fill  of  the  wall  indicate  that  the  organ  was  not  merely  dragged  through  the  sediment  but 
that  the  thin  wall  of  sediment  was  packed  in  both  horizontal  and  vertical  increments  by  the 
wall-organ. 

The  Dictyodora  animal  was  probably  a worm  or  shell-less  mollusc  which  fed  by  sediment  ingestion 
and  maintained  contact  with  the  over-lying  water  by  means  of  the  wall-organ  which  controlled 
burrow  depth  and  possibly  aided  respiration. 


OTHER  TRACES  PRESENT 
Caridolites  wilsoni  Etheridge,  Woodward  and  Jones,  1890 
Text-figs.  6,  7 

The  name  Caridolites  wilsoni  was  first  mentioned  in  Nicholson  (1873)  and  a brief  description 
appeared  in  Etheridge,  Woodward  and  Jones  (1890),  which  must  rank  as  the  type  description. 
Nicholson’s  original  (1872)  manuscript  with  a description  and  figure  of  C.  vw/som'has  been  published 
recently  together  with  a discussion  (Benton  and  Trewin  1978,  p.  10,  pi.  3)  in  which  Nicholson’s 
interpretation  that  the  trace  was  made  by  the  tail  spines  of  shoals  of  swimming  Ceratiocaris  is 
rejected. 

The  traces  are  generally  about  1 mm  wide  and  may  consist  of  a slight  central  ridge  bounded  by 
hollows  or  a single  ridge,  or  the  counterpart  of  either.  The  traces  are  generally  nearly  straight  for  from 
10-50  mm  before  disappearing  or  turning  fairly  sharply  on  a new  course.  Typical  examples  are  shown 
in  text-fig.  6 a-j  and  typical  profiles  in  text-fig.  61.  In  cross  section  the  traces  are  seen  to  be  burrows 
with  a vertical  depth  of  up  to  5 mm  and  consist  of  a basal  tunnel  with  a narrower  vertical  extension 
(text-fig.  6k).  These  traces  thus  resemble  minute  Dictyodora  without  the  meanders.  Caridolites 
frequently  covers  bedding  surfaces  with  a confusion  of  burrows  as  in  text-fig.  7. 


BENTON  AND  TREWIN:  DICTYODORA 


511 


text-fig.  6.  Caridolites  wilsoni.  a-j,  typical  burrow  traces  as 
seen  on  bedding  surfaces;  a-c,  AUGD  10675;  d-f,  AUGD 
10748;  i,  j,  AUGD  7055,  Grieston  Quarry;  g,  h,  AUGD 
10723,  Thornylee  Quarry,  k,  typical  vertical  cross  sections  of 
burrows.  /,  profiles  of  surface  expressions  of  the  burrows. 


text-fig.  7.  Caridolites  wilsoni.  Bedding 
surface  covered  with  typical  examples,  x 1 , 
AUGD  10674,  Grieston  Quarry. 


Caridolites  is  abundant  at  both  Grieston  and  Thornylee  and  is  frequently  associated  with  both 
D.  scotica  and  D.  tenuis.  It  seems  possible  that  Caridolites  represents  the  activities  of  juvenile 
Dictyodora  animals  which  had  not  developed  sufficiently  to  meander.  Certainly  the  observed  size 
ranges  of  the  traces  fit  this  possibility. 

Genus  nereites  MacLeay,  1839 

Nereites  is  rare  in  the  Thornylee-Grieston  assemblage,  with  only  two  clear  examples  of  this  surface 
trace  seen.  Sediment  surface  texture  was  probably  not  suited  to  preservation  of  surface  trails  and 
most  were  probably  removed  by  turbidity  currents.  The  slaty  muds  and  silts  generally  do  not  split  at 
the  top  surfaces  of  beds.  The  common  association  of  Nereites  surface  traces  in  sequences  with 
Dictyodora  burrows  of  similar  width  raises  the  speculation  that  Nereites  could  be  a surface  trace  of 
the  Dictyodora  animal  moving  from  one  feeding  spot  to  another. 


512 


PALAEONTOLOGY,  VOLUME  23 


CONCLUSIONS 

The  deep  water  ichnofauna  of  the  greywacke/shale  turbidite  facies  of  the  Llandovery  in  southern 
Scotland  is  dominated  by  two  species  of  Dictyodora.  The  small  burrow  Caridolites  is  probably  the 
juvenile  burrow  of  the  ‘ Dictyodora ’ animal.  Nereites  is  also  present  but  rare,  probably  owing  to 
original  sediment  texture  and  preservation. 

Crossopodia  scotica  is  shown  to  be  a Dictyodora,  and  it  is  suggested  that  it  should  be  rejected  as  the 
type  species  of  Crossopodia,  being  replaced  by  C.  lata. 

Acknowledgements.  We  thank  the  following  for  the  loan  of  specimens  and  study  facilities:  Dr.  R.  B.  Rickards, 
Sedgwick  Museum,  Cambridge;  Dr.  A.  W.  A.  Rushton,  Geological  Survey  Museum,  I.G.S.,  London;  Dr. 
W.  D.  I.  Rolfe,  Hunterian  Museum,  Glasgow;  Dr.  R.  Wilson  and  Mr.  P.  J.  Brand,  I.G.S.,  Edinburgh;  and  Mr. 
D.  N.  Lewis,  British  Museum  (Natural  History). 

REFERENCES 

benton,  m.  j.  and  trewin,  n.  h.  1978.  Discussion  and  comments  on  Nicholson’s  1872  manuscript  ‘Contributions 
to  the  study  of  errant  annelides  of  the  older  Palaeozoic  rocks’.  Pubis.  Dep.  Geol.  Miner.  Univ.  Aberdeen,  1, 
1-16. 

delgado,  j.  f.  n.  1910.  Terrains  paleozoiques  du  Portugal.  Etude  sur  les  fossiles  de  schistes  a nereites  de  San 
Domingos  et  des  schistes  a nereite  et  a graptolites  de  Barrancos.  Commis.  Serv.  geol.  Portugal,  56,  1 68, 
pis.  1-51. 

emmons,  E.  1844.  The  laconic  system.  68  pp.,  6 pis.,  Albany,  New  York. 

Etheridge,  R.,  woodward,  H.  and  jones,  T.  R.  1890.  Seventh  report  of  the  committee  on  the  fossil  Phyllopoda  of 
the  Palaeozoic  rocks.  Rep.  Br.  Ass.  Advmt.  Sci.  for  1889,  Trans.  Sections,  1890,  63-68. 
geinitz,  h.  b.  1867.  Uber  Dictyophytonl  Liebeanum  Gein.  aus  dem  Culmschiefer  vom  Heersberge  zwischen  Gera 
und  Weyda.  Neues  Jb.  Miner.  Geol.  Palaont.  1867,  286-288. 
hantzschel,  w.  1962.  Trace  fossils  and  problematica.  Pp.  W177-W245.  In  moore,  r.  g.  (ed.).  Treatise  on 
Invertebrate  Paleontology,  Part  W,  Miscellanea.  Geol.  Soc.  Am.  and  Univ.  Kansas  Press. 

— 1975.  Trace  fossils  and  problematica.  In  teichert,  c.  (ed.).  Treatise  on  Invertebrate  Paleontology,  Part  W, 
Miscellanea,  Supplement  1,  W1-W269,  Geol.  Soc.  Am.  and  Univ.  Kansas  Press. 

harkness,  R.  1855.  On  the  anthracitic  shales  and  the  fucoidal  remains  occurring  in  the  Lower  Silurian  rocks  of 
the  south  of  Scotland.  Q.  Jl  Geol.  Soc.  Lond.  11, 468-476. 
lapworth,  c.  1870.  On  the  Lower  Silurian  rocks  of  Galashiels.  Geol.  Mag.  7,  204-209,  279-284. 
m‘coy,  f.  1851a.  On  some  new  Protozoic  Annulata.  Ann.  Mag.  nat.  Hist.,  ser.  2,  7,  394-396. 

— 18516.  A systematic  description  of  the  British  Palaeozoic  fossils  in  the  Geological  Museum  of  the 
University  of  Cambridge.  In  Sedgwick,  a.  A synopsis  of  the  classification  of  the  British  Palaeozoic  rocks. 
Parker,  London  [publ.  1851-1855,  pp.  1-184  in  1851], 

macleay,  w.  s.  1839.  Note  on  the  Annelida.  Pp.  699-701.  In  Murchison,  r.  i..  The  Silurian  System,  pt.  II, 
Murray,  London. 

muller,  A.  H.  1962.  Zur  Ichnologie,  Taxiologie  und  Okologie  fossiler  Tiere.  Freiberger  ForschHft.  C.  151,  5-49. 

— 1971.  fiber  Dictyodora  liebeana  (Ichnia  invertebratorum),  ein  Beitrag  zur  Taxiologie  und  Okologie 
sedimentfressender  Endobionten.  Deutsch.  Akad.  Wiss.  Berlin,  Monatsber.  13,  136-151. 

nicholson,  H.  A.  1873.  Contributions  to  the  study  of  the  errant  annelides  of  the  older  Palaeozoic  rocks.  Proc.  R. 
Soc.  21,  288-290. 

— 1978.  Contributions  to  the  study  of  the  errant  annelides  of  the  older  Palaeozoic  rocks.  Pubis.  Dep.  Geol. 
Miner.  Univ.  Aberdeen,  1,  17-47. 

— and  etheridge,  r.,  jun.  1878-1881.  A monograph  of  the  Silurian  fossils  of  the  Girvan  district  in  Ayrshire. 
Vol.  1,  Blackwood,  Edinburgh  and  London. 

nicol,  J.  1850.  Observations  on  the  Silurian  strata  of  the  southeast  of  Scotland.  Q.  Jl  geol.  Soc.  Lond.  6,  53-65. 
peach,  B.  N.  and  horne,  J.  1899.  The  Silurian  rocks  of  Britain.  1,  Scotland,  xviii  + 749  pp.  Mem.  geol.  Surv.  U.K. 
Glasgow. 

pfeiffer,  H.  1959.  fiber  Dictyodora  liebeana  (Weiss).  Geologie,  8,  425-439. 

— 1968.  Die  Spurenfossilien  de  Kulms  (Dinants)  und  Devons  der  Frankenwalder  Querzone  (Thuringen). 
Jb.  Geol.  2,  651-717. 

raup,  d.  m.  and  seilacher,  A.  1969.  Fossil  foraging  behaviour:  computer  simulation.  Science,  N.Y.  166, 
994-995. 


BENTON  AND  TREWIN:  DICTYODORA 


513 


richter,  r.  1924.  Flachseebeobachtungen  zur  Palaontologie  und  Geologie.  IX.  Zur  Deutung  rezenter  und 
fossiler  Maander-Figuren.  Senckenbergiana,  6,  141-157. 

1928.  Psychische  Reaktionen  fossiler  Tiere.  Palaeobiologica,  1,  225-244. 

ruchholz,  k.  1967.  Zur  Ichnologie  und  Fazies  des  Devons  und  Unterkarbons  im  Harz.  Geologie,  16,  503-527. 
schimper,  w.  p.  and  schenk,  a.  1879-90.  Palaeophytologie.  In  zittel,  k.  a.  von  (ed.).  Handbuch  der 
Palaontologie,  II,  1-152  (1879). 

seilacher,  a.  1955.  Spuren  und  Fazies  im  Unterkambrium.  Pp.  86-143.  In  schindewolf,  o.  h.  and  seilacher,  a. 
Beitrage  zur  Kenntnis  des  Kambriums  in  der  Salt  Range  (Pakistan).  Akad.  fViss.  Lit.  Mainz,  math.-nat.  Kl., 
Abh.  10. 

1967.  Fossil  behaviour.  Scien.  Am.  217  (2),  72-80. 

toghill,  p.  and  strachan,  i.  1970.  The  graptolite  fauna  of  Grieston  Quarry,  near  Innerleithen,  Peeblesshire. 
Palaeontology,  13,  511-521. 

trewin,  N.  H.  1979.  Transported  graptolites  and  associated  tool  marks  from  Grieston  Quarry,  Innerleithen, 
Peeblesshire.  Scott.  J.  Geol.  15,  287-292. 

weiss,  E.  1884a.  Vorlegung  des  Dictyophytum  Liebeanum  Gein.  aus  der  Gegend  Von  Gera.  Sitz.-Ber.  Gen.  naturf. 
Freunde,  Berlin,  1884,  17. 

— 18846.  Beitrag  zur  Culm-Flora  von  Thiiringen.  Jb.  Preuss.  Geol.  Landesanst.  1883,  81-100. 
zimmermann,  E.  1889.  Uber  die  Gattung  Dictyodora.  Z.  dt.  geol.  Ges.  41,  165-167. 

1891.  Neue  Beobachtungen  an  Dictyodora.  Ibid.  43,  551-555. 

1892.  Dictyodora  liebeana  (Weiss)  und  ihre  Beziehungen  zu  Vexillum  (Rouault),  Palaeochorda  marina 

(Gein.)  und  Crossopodia  henrici  (Gein.),  Jb.  Ges.  Freunde  Naturwiss.  Gera,  32-35,  28-63. 

M.  J.  BENTON 

Department  of  Geology 
University  of  Newcastle 
Newcastle-upon-Tyne,  NE1  7RU 

N.  H.  TREWIN 

Department  of  Geology  and  Mineralogy 
Marischal  College 
University  of  Aberdeen 
Aberdeen,  AB9  IAS 


Manuscript  received  15  June  1979 

Revised  manuscript  received  3 September  1979 


LOWER  CRETACEOUS  TEREB R ATULI DAE 
FROM  SOUTH-WESTERN  MOROCCO 
AND  THEIR  BIOGEOGRAPHY 

by  FRANK  A.  MIDDLEMISS 


Abstract.  The  terebratulid  brachiopods  contained  in  the  Gentil  and  Whitaker  Collections  from  the  Lower 
Cretaceous  of  south-west  Morocco  have  been  revised.  Although  the  majority  of  the  species  are  confined  to 
south-west  Morocco,  the  affinities  of  the  fauna  are  with  the  faunas  of  the  shallow  marine  regions  bordering 
Tethys,  such  as  the  Jura  region,  eastern  Spain,  the  Crimea,  and  the  northern  Caucasus;  the  Tethyan  pygopid 
brachiopods  characteristic  of  the  Rif  in  northern  Morocco  are  almost  absent.  The  fauna  thus  constitutes  a Jura- 
type  assemblage  situated  on  the  southern  side  of  Tethys.  In  the  systematic  section  a new  genus  Paraboubeithyris 
is  erected;  also  seven  new  species:  Loriolithyris  melaitensis,  L.  marocensis,  Boubeithyris  tibourrensis , B.  pleta, 
Paraboubeithyris  plicae,  Kutchithyris  kennedyi,  and  Juralina  ecruensis.  The  genera  Kutchithyris  and  Juralina, 
previously  described  from  the  Jurassic,  are  shown  to  have  survived  into  the  Lower  Cretaceous.  Terebratula 
subsella  Leymerie  is  referred  to  Kutchithyris. 

This  paper  consists  mainly  of  a revision  of  the  terebratulids  contained  in  two  important  collections, 
the  Gentil  Collection  in  the  Collection  de  Paleontologie  of  the  Universite  Pierre  et  Marie  Curie,  Paris, 
and  the  Whitaker  Collection  in  the  British  Museum  (Natural  History),  London.  All  the  specimens 
came  from  the  Lower  Cretaceous  (Berriasian  to  Aptian  inclusive)  of  an  area  at  the  seaward  end  of  the 
High  Atlas  in  south-western  Morocco,  extending  some  40  kilometres  inland  between  Agadir  in  the 
south,  Essaouira  (Mogador)  in  the  centre,  and  Safi  in  the  north. 

Louis  Gentil,  who  was  born  at  Algiers  in  1868  and  died  in  Paris  in  1925,  was  a pioneer  in  the  study 
of  the  geology  of  Morocco.  His  first  major  contribution  was  the  exploration  of  the  Tafna  basin.  Later 
he  became  a member  of  the  Segonzac  exploratory  mission  to  the  Atlas  Mountains  and  eventually 
head  of  the  mission.  He  was  the  author  of  numerous  publications,  particularly  on  the  geology  of  the 
Atlas,  almost  up  to  the  time  of  his  death  including,  most  notably,  the  first  geological  map  of  Morocco, 
which  appeared  in  1923.  J.  J.  S.  Whitaker  was  not  a geologist  but  a Christian  missionary  who  worked 
in  Morocco  during  the  early  years  of  this  century.  His  collection  was  made  at  one  locality  only  (see 
p.  519  below)  and  very  probably  on  one  occasion.  Figured  specimens  are  in  the  British  Museum 
(Natural  History)  (BM)  or  the  Collection  de  Paleontologie,  Universite  Pierre  et  Marie  Curie,  Paris 
(Gentil  Coll.). 

THE  LOWER  CRETACEOUS  OF  SOUTH-WESTERN  MOROCCO 

The  Lower  Cretaceous  geology  of  the  area  was  described  by  Roch  (1930)  and  that  of  the  southern 
part  by  Ambroggi  (1963);  Gigout  (1951)  included  the  extreme  northern  part,  around  Safi,  in  his 
survey;  Ager  (1974)  gave  a brief  summary  in  English.  All  agree  that  south-western  Morocco  was,  in 
Lower  Cretaceous  times,  a marine  depositional  basin  opening  westwards  towards  the  ocean,  cut  off 
from  the  marine  deposits  of  the  same  age,  but  quite  different  lithofacies  and  fauna,  in  the  Rif  arc  to 
the  north  by  the  interposition  of  the  positive  block  of  the  Moroccan  Meseta  and  from  the  marine  area 
of  the  Algerian  high  plateaux  by  the  emergent  central  massif  of  the  High  Atlas.  At  each  stage  of  the 
Lower  Cretaceous  the  most  fully  marine  conditions,  presumably  indicating  the  deepest  water,  are 
found  in  the  extreme  west,  around  Cap  Ghir  and  northwards  to  the  neighbourhood  of  Cap  Tafelney. 
Passing  north-eastwards,  eastwards,  and  south-eastwards  from  this  region  one  finds  increasingly 


(Palaeontology,  Vol.  23,  Part  3, 1980,  pp.  515-556,  pis.  55-61.] 


516 


PALAEONTOLOGY,  VOLUME  23 


shallow-water  lithofacies  and  biofacies  and,  usually  within  40  or  50  kilometres,  non-marine 
deposits. 

The  deep-water  facies  around  Cap  Ghir  consists  of  green  marls  and  marly  limestones  with 
ammonites.  These  pass  eastwards  into  the  more  sandy  and  calcareous  beds,  with  brachiopod  and 
mollusc  faunas,  of  what  Roch  significantly  calls  a ‘jurassian  facies’.  These  pass  eventually  into  sub- 
continental red  beds.  The  lithological  succession  differs  markedly  from  the  monotonous  lithofacies 
of  the  ‘bathyal’  Lower  Cretaceous,  seen  in  the  Rif  and  the  Betic  region,  and  has  a general  resemblance 
to  the  successions  seen  in  the  Pre-Betic  zone  of  Spain,  north-east  Spain  (Sitges),  east-central  Sardinia, 
Provence,  and  Portugal.  It  exhibits  a very  striking  difference  from  these,  however,  in  the  absence  of 
the  massive  urgonian  limestones,  which  are  characteristically  developed  in  the  Barremian  and  Aptian 
of  those  regions,  and  of  the  rudists.  In  these  respects,  the  south-west  Moroccan  succession  is  most 
comparable  to  the  Lower  Cretaceous  of  central  Texas  and  parts  of  Coahuila  (Mexico).  The  Aptian, 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


517 


as  in  northern  Spain  and  England,  is  transgressive,  the  Gargasian  overlapping  the  earlier  divisions  on 
to  the  flanks  of  the  High  Atlas.  To  the  south  lies  the  coastal  Cretaceous  basin  of  Tarfaya,  at  first  sight 
similar  in  situation  to  the  Agadir-Essaouira  basin,  but  here  the  earlier  part  of  the  Cretaceous  is  non- 
marine, marine  sedimentation  starting  only  with  the  Apto-Albian  (Choubert  et  al.  1967). 


PALAEOBIOGEOGRAPHICAL  RELATIONSHIPS  OF  THE 
TEREBRATULID  FAUNA 

Endemicity.  The  fauna  contains  a high  proportion  of  endemic  species:  of  the  eleven  species  described 
eight  are  new  and  seven  of  these  are  so  far  known  only  from  south-west  Morocco.  This  is  not  unusual. 
The  terebratulids  tend  to  produce  local,  allopatric  species.  For  example,  of  the  sixteen  terebratulid 
species  in  the  English  Aptian  thirteen  are  known  only  in  south  and  south-central  England,  of  which 
three  occur  at  one  locality  only  (Middlemiss  1959).  I have  recently  (Middlemiss  1979)  pointed  to  the 
contrast  between  such  local  species  and  widespread  species  such  as  (in  the  Moroccan  fauna) 
Loriolithyris  valdensis  and  suggested  that  these  differences  were  probably  due  to  differing  lengths  of 
the  free-swimming  larval  stage.  Evidence  for  the  palaeobiogeographical  relationships  of  the  fauna 
comes  mainly  from  the  occurrence  elsewhere  of  the  widespread  species  but  also  from  the  taxonomic 
relationships  of  the  local  species. 

Loriolithyris.  L.  valdensis  is  the  most  widespread  species  of  this  genus,  occurring  in  the  Lower  Cretaceous  of 
eastern  Spain  (and  the  Balearic  Islands),  Sardinia,  southern  France,  the  Jura,  south-east  Paris  Basin,  north-east 
Bulgaria,  the  Crimea,  northern  Caucasus,  Kopet  Daga,  and  perhaps  Algeria.  L.  russillensis  shares  the  western 
part  of  this  distribution— eastern  Spain,  the  Balearic  Islands,  southern  France,  the  Jura,  and  south-east  Paris 
Basin.  L.  melaitensis  and  L.  marocensis  are  local  offshoots  from  the  stock,  not  at  present  known  outside  the 
south-west  Moroccan  basin. 


text-fig.  2.  Palaeobiogeographical  relationships  of  the  Lower  Cretaceous  terebratulids  of  south-west  Morocco. 
Distribution  of  south-west  Moroccan  Lower  Cretaceous  species  which  occur  elsewhere:  ■ Loriolithyris 
russillensis,  □ Loriolithyris  valdensis,  ♦ Cyrtothyris  middlemissi,  O Kutchithyris  kennedyi.  Distribution  of  other 
Lower  Cretaceous  species  of  Cyrtothyris : + . Distribution  of  Aptian-Cenomanian  species  of  Boubeithyris:  ☆ . 
Generalized  occurrence  of  Kutchithyris  subsella  in  the  Upper  Jurassic  and  Lower  Cretaceous:  * . Generalized 
occurrence  of  Jurassic  species  of  Juralina:  + . Generalized  boundary  of  the  Tethyan  pygopid  fauna  shown  by 

diagonal  shading. 


518 


PALAEONTOLOGY,  VOLUME  23 


Boubeithyris  and  Paraboubeithyris.  The  three  species  here  ascribed  to  these  genera  are  all  local  to  south-west 
Morocco  but  the  genus  Boubeithyris,  of  which  Paraboubeithyris  is  perhaps  a specialized  development,  is 
represented  by  a species  in  the  Aptian  of  the  Jura,  by  two  species  in  the  Albian  of  England  and  by  one  species  in 
the  Cenomanian  of  Belgium  and  western  France. 

Cyrtothyris.  C.  middlemissi,  the  south-west  Moroccan  species,  is  known  also  in  eastern  Spain  and  southern 
France.  The  genus  is  more  widespread,  being  represented  by  species  in  the  early  Cretaceous  of  north  Germany, 
north-east  England,  and  east  Greenland  and  the  Aptian  of  the  Jura  and  southern  France.  Imlay’s  species 
Terebratula  sillimani  and  T.  tamaulipana  (Imlay  1937),  from  the  Valanginian-Hauterivian  of  northern  Mexico, 
probably  belong  to  this  genus. 

Kutchithyris.  K.  brivesi  is  a highly  distinctive  form  confined  to  south-west  Morocco  but  K.  kennedyi  is  known  also 
in  the  Lower  Cretaceous  of  eastern  Spain,  the  Balearic  Islands,  and  southern  France,  the  southern  part  of  the 
same  distribution  area  as  L.  russillensis.  Other  species  of  the  genus  are  found  in  the  Middle  Jurassic  of  India  and, 
according  to  Buckman  (1918),  Europe.  I here  refer  Terebratula subsella  Leymerie  to  this  genus.  This  species  has  a 
widespread  occurrence  in  the  Upper  Jurassic  of  Europe  and  is  known  (but  undescribed)  in  the  Lower  Cretaceous 
of  eastern  Spain. 

Juralina.  This  genus,  as  interpreted  by  recent  authors  (especially  Boullier  1976),  occurs  in  the  Upper  Jurassic  of  a 
wide  area  of  Europe  north  of  the  Alps  from  England  to  Russia  and  also  of  Crete  ( J . immanis—see  Bonneau, 
Beauvais,  and  Middlemiss  1975)  and  Sicily  (Boullier  1976).  J.  ecruensis  is  the  first  species  of  the  genus  to  be 
described  from  the  Cretaceous. 

Discussion.  The  Lower  Cretaceous  terebratulids  of  Europe  can  be  divided  into  three  geographical 
faunas:  the  boreal  fauna  in  the  north,  the  Tethyan  fauna  with  its  distinctive  Pygopinae,  and  between 
them  the  Jura  fauna.  The  last  is  so  named  after  the  area  in  which  the  fauna  is  richest  and  best  known, 
but  the  character  of  the  Jura  fauna  is  essentially  that  of  a neritic  assemblage  occupying  an  optimum 
situation  on  the  border  of  the  deeper-water  Tethyan  region  and  extending  approximately  parallel  to 
the  border  of  Tethys  from  the  Iberian  Peninsula  eastwards  to  Turkmenistan.  In  this  sense,  the  Lower 
Cretaceous  fauna  of  south-west  Morocco  falls  into  place  as  an  extension  of  the  Jura  fauna  to  the 
south  of  the  Tethyan  fauna  which  is  so  strongly  developed  in  the  Rif. 

The  affinities  of  our  terebratulids  are  essentially  with  the  Jura  brachiopod  fauna.  This  is  generally 
true  of  the  cephalopods  listed  and  figured  by  Roch,  Ambroggi,  and  Gigout.  Characteristic  Tethyan 
genera  such  as  Lytoceras  (Valanginian-Hauterivian),  Phylloceras  (Hauterivian),  Desmoceras 
(Barremian),  Pulchellia  (Barremian),  Duvalia  (Valanginian),  Hibolites  (Valanginian)  occur  but  are 
almost  confined  to  the  deep-water  region  of  the  extreme  west.  Further  east  the  cephalopods  are  noted 
by  Roch  as  being  of  ‘Jura  type’  and  include  such  genera  as  Acanthodiscus  and  Leopoldia.  There  is 
scarcely  a trace  in  the  pre-Aptian  Cretaceous  of  the  Tethyan  pygopines  which  characterize  the  Rif 
and  the  Betic  region  (Geyssant  1966).  The  ‘jurassian’  affinities  of  the  faunal  facies  were  clearly 
recognized  by  Roch  and  Gignoux  (1955).  Ager  (1974)  has  recorded  the  discovery  of  Nucleata  cf. 
jacobi  in  the  Aptian  or  Albian  near  Tamzargout.  This  seems  to  be  the  only  recorded  occurrence  of 
pygopine  brachiopods  in  the  Lower  Cretaceous  of  south-west  Morocco— a feeble  sign  of  southward 
Tethyan  spread’  simultaneous  with  those  transgressions  which  were  causing  northward  movement 
of  southern  species  into  north  Spain,  England,  and  north  Germany  (Middlemiss  1979).  The  specimen 
from  Safi  figured  by  Gigout  (1951,  pi.  9,  figs.  35-38)  as  T.  euthymi  is  a terebratellidine  related  to 
‘ Terebratula ’ moreana  d’Orbigny. 

Kutchithyris,  in  the  Lower  Cretaceous,  does  not  occur  north  of  southernmost  France  and  is  one  of 
those  sub-Tethyan  forms  (Middlemiss  1979)  which  are  sensitive  indicators  of  the  advance  and  retreat 
of  the  Tethyan  fauna.  K.  subsella  shows  this  well.  In  the  Oxfordian,  a period  of  major  expansion  of 
the  Tethyan  fauna  (Arkell  1956),  it  is  found  throughout  a large  part  of  central  Europe— England, 
northern  France,  northern  and  south-western  Germany,  southern  Poland,  the  Russian  Platform.  By 
Kimmeridgian  times  it  extended  no  further  north  than  the  Boulonnais.  The  Volgian  saw  a further 
southward  retreat  to  the  Pays  de  Bray,  its  place  in  England  and  the  Boulonnais  being  taken  by  boreal 
forms.  In  the  Lower  Cretaceous  it  has  so  far  been  found  only  in  the  Pre-Betic  region  of  Spain,  on  the 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


519 


margin  of  Tethys.  Juralina  may  also  be  a sub-Tethyan  genus  whose  history  is  possibly  similar  to  that 
of  K.  subsella. 

Reconstruction  of  plate  positions  as  they  were  in  Lower  Cretaceous  times  shows  the  area  of  the 
Jura  faunas  as  much  more  linear  than  it  is  now.  Provence,  eastern  Spain,  Sardinia,  the  Balearic 
Islands,  and  south-west  Morocco  form  a linear  belt  which,  extended  westwards,  would  include  the 
western  Gulf  region  of  the  U.S.A.  and  the  northern  parts  of  Mexico.  The  neritic  Lower  Cretaceous  of 
these  latter  regions  is  in  this  sense  an  extension  of  the  area  of  the  Jura  fauna.  Unfortunately 
brachiopods  are  rare  but  Imlay  (1940)  remarked  of  the  Neocomian  faunal  assemblage  of  northern 
Mexico  that  it  was  remarkably  similar  to  that  of  France,  England,  and  Switzerland  and  belonged 
decidedly  to  the  ‘Mediterranean’  province.  His  species  T.  coahuilensis  is  certainly  close  to  and 
probably  synonymous  with  Sellithyris  carteroniana  d’Orbigny,  one  of  the  most  characteristic  Jura 
species.  It  seems  a reasonable  forecast  that  neritic  Lower  Cretaceous  brachiopod  assemblages  of 
‘Jura  fauna’  affinities  will  some  day  be  found  in  the  south-eastern  or  Gulf  continental  shelf  deposits  of 
the  U.S.A.  or  the  north-western  continental  shelf  deposits  of  Africa.  Unfortunately  those  of  the 
offshore  part  of  the  Tarfaya  basin  have  yielded  no  brachiopods. 


STRATIGRAPHIC  AGES  OF  SPECIMENS  IN  THE  WHITAKER  AND 
GENTIL  COLLECTIONS 

Whitaker  left  no  record  of  the  age  of  the  strata  from  which  he  made  his  collection  and  it  has  not  so  far 
proved  possible  to  trace  the  exact  locality.  All  the  specimens  were  obtained  from  one  locality, 
recorded  as:  ‘Ecru,  Mogador,  Morocco.  500  ft.  on  plateau  edge  of  1000  ft.  elevation’.  The  age  can 
only  be  assessed  on  the  internal  evidence  of  the  fauna  and  appears  to  be  either  Hauterivian  or 
Barremian.  The  species  represented  all  occur  elsewhere  in  south-west  Morocco  in  both  the 
Hauterivian  and  the  Barremian,  whereas  not  all  occur  in  the  Yalanginian  or  Aptian. 

Four  species  are  represented  in  the  Whitaker  Collection,  in  the  following  numbers:  Loriolithyris 
russillensis,  57;  L.  valdensis,  39;  Juralina  ecruensis,  46;  Kutchithyris  kennedyi,  1 . The  predominance  of 
L.  russillensis  would  suggest,  on  analogy  with  the  occurrence  of  the  species  in  Switzerland  and 
France,  a Barremian  age.  The  distribution  of  these  four  species  in  the  Gentil  Collection  is  as  follows: 


L.  russillensis 

Hauterivian 

7 

J.  ecruensis 

Valanginian 

25 

Barremian 

39 

Hauterivian 

2 

Aptian 

9 

Barremian 

15 

L.  valdensis 

Valanginian 

25 

K.  kennedyi 

Hauterivian 

2 

Hauterivian 

63 

Barremian 

1 

Barremian 

104 

Aptian 

16 

In  general  these  statistics  again  support  a Barremian  age  for  the  Whitaker  Collection  but  they  may 
reflect  nothing  more  than  the  accidents  of  collection. 

I have  followed  stratigraphic  ages  given  on  the  labels  of  the  Gentil  Collection  because  it  was  not 
possible  to  check  each  locality  in  the  field,  but  there  are  some  arguments  supporting  the  general 
validity  of  these  labels,  even  though  there  must  be  a number  which  are  wrong.  Analysis  of  all  the 
localities  given  on  the  labels  shows  that  all  the  specimens  from  any  one  locality  are  assigned 
consistently  either  to  a single  stage  or  to  two,  or  rarely  three,  adjacent  stages.  Thus  a logical  series  of 
localities  can  be  set  out,  ranging  from  those  credited  with  yielding  only  Berriasian  and  Valanginian 
fossils  to  those  credited  with  yielding  fossils  only  of  Clansayesian  age. 


520 


PALAEONTOLOGY,  VOLUME  23 


SYSTEMATIC  PALAEONTOLOGY 

Order  terebratulida  Waagen,  1883 
Suborder  terebratulidina  Waagen,  1883 
Superfamily  terebratulacea  Gray,  1 840 
Family  terebratulidae  Gray,  1840 
Subfamily  sellithyridinae  Muir-Wood,  1965 

Remarks.  Loriolithyris  and  Boubeithyris  are  closely  related  sellithyridine  genera.  The  corniced  hinge 
plates  which  are  the  most  distinguishing  feature  of  Boubeithyris  are  essentially  the  same  in  detailed 
structure  as  the  piped  hinge  plates  of  Loriolithyris.  Both  genera  essentially  have  small  crural  bases 
(attached  to  the  inner  edges  of  the  hinge  plates)  which  become  encased  in  successive  layers  of 
secondary  skeletal  tissue  (PI.  60,  fig.  2;  PI.  61,  figs.  2,  3).  The  function  of  this  is  presumably  to 
strengthen  the  junction  of  hinge  plates  and  crural  bases.  These  structures  are  not  inner  hinge  plates, 
which  some  authors  claim  to  be  present  in  Terebratula,  although  Muir-Wood  (1965,  p.  H775)  denies 
their  presence  in  that  genus,  because  they  show  no  sign  of  having  taken  part  in  any  way  in  the 
attachment  of  the  dorsal  pedicle  muscles.  Boubeithyris  and  Loriolithyris  differ  mainly  in  the  shape  of 
the  hinge  plates— concave  and  corniced  in  Boubeithyris , concave  to  sigmoid  and  piped  in 
Loriolithyris.  Externally  Boubeithyris  is  distinguished  especially  by  the  close  spacing  of  the  plicae  of 
the  anterior  commissure.  Both  differ  from  Sellithyris  in  having  accessory  structures  (cornicing  or 
piping)  on  the  hinge  plates  and  in  their  much  less  pentagonal  external  form. 

Paraboubeithyris  has  an  internal  structure  which  is  closely  related  to  that  of  Boubeithyris. 
Externally  P.  plicae  looks  different  at  first  sight  from  Boubeithyris  spp.  but  similarities  include  the 
convex  cardinal  slopes,  small  size  of  the  median  sinus,  and  the  late  development  of  folding.  The 
external  differences,  however,  seem  too  great  to  allow  the  species  to  be  included  in  Boubeithyris. 
P.  plicae  is  here  regarded  as  a specialized  local  offshoot  from  the  Boubeithyris  stock. 

Genus  loriolithyris  Middlemiss,  1968 
Type  species.  Terebratula  russillensis  de  Loriol,  1866. 

Species  included.  T.  russillensis  de  Loriol,  T.  valdensis  de  Loriol,  L.  melaitensis  nov.,  L.  marocensis  nov.  Range: 
Berriasian  to  Aptian. 


explanation  of  plate  55 

Figs.  1-4.  Loriolithyris  russillensis  (de  Loriol).  Whitaker  Coll.  \a-d,  typical  form,  plaster  cast  of  specimen 
sectioned  (see  text-fig.  5),  BM  BB  76544.  2 a-c,  wide  latifrons-  like  form,  plaster  cast  of  specimen  sectioned  (see 
text-fig.  7),  BM  BB  76552.  3 a-d,  small  sharply  folded  form,  plaster  cast  of  specimen  sectioned  (see  text-fig.  6) 
BM  BB  76543.  4 a-d,  thick  latifrons- like  form,  BM  B 17293. 

Figs.  5-9.  Loriolithyris  valdensis  (de  Loriol).  5 a-c,  typical  form,  plaster  cast  of  specimen  sectioned  (see  text- 
fig.  11),  BM  BB  76545.  Whitaker  Coll.  6 a-d,  juvenile  rectimarginate  form,  BM  BB  76546,  Whitaker  Coll. 
la-d , juvenile  incipiently  biplicate  form  BM  BB  76549,  Whitaker  Coll.  8 a-d,  elongate  adult  form,  BM  BB 
76554,  Whitaker  Coll.  9 a-d,  wide  adult  form,  S. 546/1/12,  Gentil  Coll.,  Upper  Hauterivian,  loc.  unknown. 

Fig.  10 a-c.  Loriolithyris  melaitensis  sp.  nov.  Plaster  cast  of  specimen  sectioned  (see  text-fig.  12),  S.556/1, 
Gentil  Coll.,  Hauterivian,  Tizi  Ouarioum. 

Figs.  11  a-d.  Loriolithyris  melaitensis  sp.  nov.  Holotype,  S. 556/2,  Gentil  Coll.,  Barremian,  Ait  Ben  Melait, 
Ida  ou  Guelluill. 


All  natural  size. 


PLATE  55 


middlemiss,  Cretaceous  Terebratulidae 


522 


PALAEONTOLOGY,  VOLUME  23 


Loriolithyris  russillensis  (de  Loriol) 

Plate  55,  figs.  1-4;  text-figs.  3-7 

* 1866  Terebratula  russillensis  de  Loriol,  p.  88,  pi.  E,  figs.  12-15. 

1867  Terebratula  russillensis  de  Loriol,  p.  393,  pi.  C,  figs.  28-31. 

1869  Terebratula  russillensis  de  Loriol,  p.  28,  pi.  4,  fig.  1 . 
vl872  Terebratula  russillensis  de  Loriol;  Pictet,  p.  68,  pi.  202,  figs.  1-8. 
vl872  Terebratula  latifrons  Pictet,  p.  67,  pi.  201,  figs.  16-17. 

71964  Sellithyris  (7)russillensis  (de  Loriol);  Ager,  p.  340. 
non  1966  Sellithyris  russillensis  (de  Loriol);  Bogdanova  and  Lobacheva,  p.  53,  pi.  5,  figs.  5-6. 
vl968  Loriolithyris  russillensis  (de  Loriol);  Middlemiss,  p.  176,  pi.  A,  figs.  1-4. 

Lectotype.  Museum  d’Histoire  Naturelle,  Geneva  (Pictet  Collection),  no.  CB  1520.  Designated  Middlemiss 
1968.  Fig.  Pictet  and  de  Loriol  1872,  pi.  202,  fig.  4;  from  the  urgonian  of  La  Russille,  Yaud,  Switzerland. 

Material.  Fifty-seven  specimens  from  the  Whitaker  Collection.  About  fifty-five  specimens  in  the  Gentil 
Collection. 

Remarks.  Specimens  from  Morocco  tend  to  be  wider  and  thinner,  in  relation  to  length,  than  the 
typical  members  of  the  species  from  La  Russille  and  Orgon  and  many  have  the  characters  of  the  form 
described  by  Pictet  (1872)  as  Terebratula  latifrons.  I have  previously  (Middlemiss  1968a)  believed  the 
latter  form  to  be  a variety  of  Loriolithyris  russillensis  and  experience  of  the  Moroccan  fauna  has 
reinforced  this  belief.  Forms  from  the  Jura  region  which  Pictet  recognized  as  T.  latifrons  (Geneva 
Museum)  are  distinct  because  of  their  decidedly  small  umbones  and  foramina,  not  because  of  their 
wide  depressed  shape.  They  usually  display  well-developed  russillensis-like  folding  of  the  shell  and  as 
regards  shape  there  seems  to  be  a complete  gradation  between  the  two  species.  In  both  south-west 
France  and  south-west  Morocco  forms  apparently  referable  to  L.  russillensis  show  continuous 
variation,  in  the  same  assemblages,  into  other  forms  with  the  same  characters  except  for  the 
proportions  of  shell  shape,  which  are  those  of  T.  latifrons.  The  forms  with  decidedly  small  umbones 
and  foramina  do  not  occur  in  these  regions.  The  internal  skeletal  arrangements  revealed  by  serial 
sectioning  are  the  same  in  all  these  forms:  the  concave  piped  hinge  plates,  situated  close  to  the  floor  of 
the  brachial  valve,  and  the  sigmoid  passage  from  inner  socket  ridge  to  hinge  plate,  are  unmistakeable. 
Pictet  records  his  typical  T.  latifrons  forms  only  from  the  Upper  Valanginian  of  Villers-le-Lac  and 
Vesency.  L.  russillensis  was  apparently  a species-group  very  variable  in  proportions  of  length,  width, 
and  thickness,  some  members  of  which,  in  part  of  the  Jura  region  and  for  a short  time  in  the  Upper 
Valanginian,  became  locally  sufficiently  differentiated  to  deserve  recognition  as  a subspecies 
Latifrons' . 


30 


20 


text-fig.  3.  Scatter  diagrams  of  relationships  of  width  to  length  and  thickness  to  length  in  Loriolithyris 
russillensis  from  the  Whitaker  Collection. 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


523 


n 

10 

9 


u 7 
O 

ffl  6 
c 

< 5 

4 

3 

2 


1 2 3 4 5 6 7 8 9 10  11  12  13  14  15  16 

Posterior 


5 


15  20  25  30 

Length 


text-fig.  4.  Scatter  diagrams  of  the  posterior/anterior  ratio  in  Loriolithyris  russillensis  from  the  Whitaker  Coll. 


The  main  differences  between  this  species  and  L.  valdensis  are  that  L.  valdensis  is  longer  is  relation 
to  both  width  and  thickness  and  has  a higher  P/A  ratio  than  L.  russillensis.  These  points  are 
graphically  illustrated,  as  far  as  the  Moroccan  specimens  are  concerned,  in  text-figs.  3, 4, 8, 9,  and  10. 
Internally,  a point  of  distinction  is  that  in  L.  russillensis  the  hinge  plates  are  close  to,  or  even  in 
part  in  contact  with,  the  floor  of  the  brachial  valve,  whereas  in  L.  valdensis  they  are  raised  clearly 
above  the  floor  of  the  valve  for  their  whole  width.  It  can  be  added  that,  internally,  L.  russillensis 
has  a very  short  loop,  little  more  than  1 mm  from  the  crural  processes  to  the  transverse  band  in 
adult  shells.  Unfortunately  it  is  characteristic  of  species  of  Loriolithyris  that  the  transverse  band 
is  delicate  and  seldom  preserved  and  I have  never  yet  seen  this  structure  in  L.  valdensis. 


text-fig.  5.  Transverse  sections  through  a small,  strongly  folded  specimen  of  Loriolithyris  russillensis.  Sections 
1.8  and  2.0  are  enlarged  in  order  to  show  the  shape  of  the  juvenile  hinge  plates  enclosed  within  the  cardinal 
process  (punctate  tissue  is  stippled  in  section  1 .8).  Section  4.2  is  enlarged  in  order  to  show  the  structure  of  the 
piped  hinge  plates.  BM  BB  76544,  Whitaker  Coll.  A — scale  for  sections,  1.8,  2.0  and  4.2.  B -scale  for  the 

remaining  sections. 


524 


PALAEONTOLOGY,  VOLUME  23 


Distribution.  Ager  (1964)  claims  this  species  in  the  Berriasian  of  the  southern  Jura  and  Pictet  (1872) 
notes  it  in  the  Valanginian  of  Sainte-Croix  (Vaud).  It  certainly  occurs  in  the  Hauterivian  of  Vaud, 
Doubs,  Haute-Marne,  and  Yonne  and  of  Les  Corbieres  (Aude).  It  is  at  its  most  abundant,  however, 
in  the  Barremian  of  Vaud,  Jura,  the  south-east  Paris  Basin,  Bouches-du-Rhone,  Gard,  Aude,  eastern 
Spain,  and  Ibiza.  It  occurs  very  rarely  in  the  Aptian  of  Aude.  In  south-west  Morocco  it  ranges  from 
the  Hauterivian  to  Aptian  inclusive. 


text-fig.  6.  Transverse  sections  through  a small,  strongly  folded  specimen  of  Loriolithyris  russillensis  to  show 
the  short  loop.  Section  1.8  is  enlarged  in  order  to  show  the  shape  of  the  juvenile  hinge  plates  enclosed  in  the 
cardinal  process.  Sections  2.2  and  2.6  are  enlarged  in  order  to  show  the  primary  hinge  plates  (stippled).  The 
maximum  height  of  the  crural  processes  is  seen  in  section  3.4.  BM  BB  76543,  Whitaker  Coll.  A— scale  for 
sections  1.8,  2.2,  and  2.6.  B— scale  for  the  remaining  sections. 


Loriolithyris  valdensis  (de  Loriol) 


Plate  55,  figs.  5-9;  text-figs.  8-11 


v*1868 
vl872 
non  1939 
1960 
pars  1966 
vl968 
1972 
v!975 


Terebratula  valdensis  de  Loriol,  p.  52,  pi.  4,  figs.  9-12. 

Terebratula  valdensis  de  Loriol;  Pictet,  p.  66,  pi.  201,  figs.  11-15. 

Terebratula  valdensis  var.  kentugajensis  Moisseev,  p.  200,  pi.  2,  fig.  6. 

‘ Terebratula ’ valdensis  de  Loriol;  Smirnova,  p.  374,  pi.  1,  fig.  1. 

Sellithyris  valdensis  (de  Loriol);  Bogdanova  and  Lobacheva,  p.  55,  pi.  5,  fig.  7 ( non  pi.  7,  fig.  11). 
Loriolithyris  valdensis  (de  Loriol);  Middlemiss,  p.  182,  pi.  A,  fig.  5. 

Sellithyris  valdensis  (de  Loriol);  Smirnova,  p.  81,  pi.  7,  fig.  5. 

Loriolithyris  valdensis  (de  Loriol);  Dieni  and  Middlemiss,  p.  182,  pi.  36,  figs.  9-10. 


Lectotype.  Museum  d’Histoire  Naturelle,  Geneva  (Arzier  Collection),  no.  CB  1505.  Designated  Middlemiss 
1968.  Fig.  de  Loriol  1868,  pi.  4,  figs.  9 a-d,  from  Bed  B,  Valanginian,  Arzier  Quarry,  Vaud,  Switzerland. 


Material.  Thirty-nine  specimens  in  the  Whitaker  Collection.  About  200  specimens  in  the  Gentil  Collection. 
Eight  specimens  from  Barremian  or  Aptian,  Tizi  ou  Elma,  Agadir  (D.V.  Ager  Collection). 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


525 


text-fig.  7.  Transverse  sections  through  a broad,  latifrons- like  specimen  of  Loriolithyris  russillensis.  Sections  2.8 
and  3.2  are  enlarged  in  order  to  show  the  shape  of  the  juvenile  hinge  plates.  The  structure  of  the  piped  inner 
margin  of  the  hinge  plate  is  enlarged  at  section  4.6  (see  plate  60,  fig.  5).  The  transverse  band  was  not  preserved  in 
this  specimen.  BM  BB  76552,  Whitaker  Coll.  A— scale  for  sections  2.8,  3.2,  and  4.6  (inset).  B— scale  for  the 

remaining  sections. 


Description.  Text-figs.  8 and  9 compare  the  thirty-nine  specimens  in  the  Whitaker  Collection  with  a collection  of 
227  specimens  made  at  the  type  locality  of  Arzier  by  Monsieur  Roessinger  and  preserved  at  the  Geneva  Natural 
History  Museum.  The  isometric  development  of  length  and  width  is  well  shown  in  text-fig.  9.  Thickness  in 
relation  to  length  develops  allometrically,  although  with  a very  small  differential  growth  ratio  (text-fig.  8).  Text- 
fig.  10  shows  that  the  P/A  ratio  develops  allometrically  with  a very  wide  range  of  variation  (about  double  the 
width  of  that  shown  by  Sellithyris  sella  from  the  Isle  of  Wight  Aptian  (Middlemiss  1968ft,  fig.  9)).  The  smallest 
shells,  less  than  5 mm  in  length,  are  subcircular  in  ventral  profile  but  posterior  length  increases  allometrically 
with  growth,  at  the  expense  of  anterior  length.  There  is  a marked  tendency  for  Moroccan  specimens  to  have  a 
lower  P/A  ratio,  i.e.  to  have  a relatively  greater  anterior  length  than  those  from  the  type  area;  in  this  respect  the 
lectotype  has  an  anomalous  position. 

The  anterior  commissure  remains  rectimarginate  until  the  shell  is  about  12  mm  in  length.  It  then  passes 
through  a well-marked  uniplicate  stage  until  the  shell  reaches  a length  of  about  16  mm,  after  which  plicae  and 
sinuses  are  rapidly  developed,  shells  from  17  mm  upwards  being  normally  sulciplicate.  The  episulcate  stage  is 
occasionally  seen  at  Arzier  but  is  very  rare  in  Morocco. 

Remarks.  Differences  between  this  species  and  L.  russillensis  were  discussed  above.  Roch  remarks  on 
the  abundance  of  this  species  in  the  Valanginian  and  Barremian  of  south-west  Morocco,  especially  in 
the  Barremian  of  Jebel  Graa  and  Aghbalou. 


526 


PALAEONTOLOGY,  VOLUME  23 


L.valdensis 

Lectotype  of  L.valdensis 
K.kennedyi 


=i:Sr 


. * 


\ %.>• 


5 10  15  20  25  30 

Length 

text-fig.  8.  Scatter  diagrams  of  the  relationship  of  thickness  to  length  in  Loriolithyris  valdensis  (Arzier  and 
Whitaker  Colls.)  and  Kutchithyris  kennedyi  (all  available  specimens). 


text-fig.  9.  Scatter  diagram  of  the  relationship  of  width  to 
length  in  Loriolithyris  valdensis  (Arzier  and  Whitaker  Colls.). 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


527 


Posterior  Length 


text-fig.  10.  Scatter  diagrams  of  the  posterior/anterior  ratio  in  Loriolithyris  valdensis  from  Arzier. 


35 


Distribution.  Berriasian  and  Valanginian  of  Vaud  and  Haute-Savoie;  Valanginian  and  Hauterivian  of 
the  south-east  Paris  Basin;  Valanginian  of  Georgia  and  Hauterivian  of  the  northern  Caucasus 
(Smirnova  1972);  Neocomian  of  the  Kopet  Daga  (Bogdanova  and  Lobacheva  1966);  Hauterivian  of 
north-east  Bulgaria.  Valanginian  and  Hauterivian  of  eastern  Spain;  Barremian  of  Basses- Alpes  and 
Alpes-Maritimes.  Aptian  of  La  Presta  (Neuchatel).  In  south-west  Morocco  the  range  is  Valanginian 
to  Aptian  inclusive. 


text-fig.  1 1 . Transverse  sections  through  Loriolithyris  valdensis.  Sections  2.8-4.4  are  enlarged  in  order  to  show 
the  shape  of  the  juvenile  hinge  plates  enclosed  within  the  cardinal  process  and  the  structure  of  the  crural  bases 
within  the  piped  inner  margins  of  the  hinge  plates.  Maximum  development  of  the  crural  processes  is  seen  in 
section  7.2.  The  transverse  band  was  not  preserved  in  this  specimen.  BM  BB  76545,  Whitaker  Coll.  A— scale  for 
sections  2.8-4.4.  B— scale  for  the  remaining  sections. 


528 


PALAEONTOLOGY,  VOLUME  23 
Loriolithyris  melaitensis  sp.  nov. 


Plate  55,  figs.  10,  11;  text-fig.  12 

vl951  Terebratula  salevensis  de  Loriol;  Gigout,  p.  360,  pi.  9,  figs.  15-18. 

Types.  Holotype,  Gentil  Collection  specimen  no.  S. 556/2,  from  the  Barremian  of  Ait  Ben  Melait.  Dimensions: 
L 31,  W 28-5,  T 18-5.  Paratype,  Gentil  Collection  specimen  no.  S. 556/1  (locality  as  holotype). 

Material.  Ten  specimens  in  the  Gentil  Collection;  nine  from  the  Hauterivian  of  Tizi  Ouarioum,  one  from  the 
Barremian  of  Ait  Ben  Melait,  Ida  ou  Guelluill. 

Diagnosis.  Loriolithyris  of  elongate  oval  ventral  profile,  becoming  thick  in  adult  stage  (thickness  nearly  equal  to 
width);  P/A  ratio  slightly  more  than  1 . Valves  equally  convex.  Umbo  suberect.  Foramen  mesothyrid,  attrite, 
slightly  labiate.  Beak  ridges  rounded.  Symphytium  very  short,  but  visible.  Lateral  commissure  strongly  arched; 
anterior  commissure  sulciplicate.  Shell  not  folded  except  at  extreme  anterior.  Small  pedicle  collar  present.  Hinge 
plates  concave,  piped.  Crural  bases  well  developed.  Crural  processes  slightly  incurved.  Transverse  band  high- 
arched,  rounded. 

Remarks.  The  thick,  well-filled  appearance  of  the  shell,  the  arched  lateral  commissure,  and  the 
relative  lack  of  folding  give  this  species  a superficial  resemblance  to  Tropeothyris  salevensis 
(de  Loriol)  and  it  is  likely  that  Ambroggi’s  (1963)  record  of  T.  salevensis  in  both  Lower  and  Upper 
Barremian  of  south-west  Morocco  refers  to  this  species. 


text-fig.  12.  Transverse  sections  through  Loriolithyris  melaitensis.  Section  4.8  is  enlarged  in  order  to  show  the 
shape  of  the  juvenile  hinge  plates  enclosed  within  the  cardinal  process  and  the  boundary  between  punctate  tissue 
(stippled)  and  impunctate  laminated  tissue.  Section  5.2  is  enlarged  in  order  to  show  the  primary  hinge  plates 
(stippled).  The  crural  bases,  unusually  large  for  Loriolithyris,  are  well  shown  in  sections  6.4-7. 6.  Section  9.6 
shows  the  maximum  development  of  the  crural  processes.  S.556/1,  Gentil  Coll.,  Hauterivian,  Tizi  Ouarioum. 

A— scale  for  sections  4.8  and  5.2.  B— scale  for  the  remaining  sections. 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


529 


It  is  distinguished  from  other  species  of  Loriolithyris  especially  by  the  unusually  large  size  of  the 
crural  bases  attached  to  the  inner  edges  of  the  hinge  plates  (text-fig.  12),  but  also  by  its  external 
appearance. 

Distribution.  Hauterivian  and  Barremian  of  south-west  Morocco. 

Loriolithyris  marocensis  sp.  nov. 

Plate  56,  figs.  1,  2;  text-fig.  13 

Types.  Holotype,  Gentil  Collection  specimen  no.  S. 547/2;  age  given  as  Upper  Hauterivian  (locality  unknown). 
Dimensions:  L 49-75,  W 32,  T 26-25.  Paratype,  Gentil  Collection  specimen  no.  S. 547/1. 

Material.  Sixteen  specimens  in  the  Gentil  Collection:  four  from  the  Hauterivian  (including  Oued  Tidzi),  two 
from  the  Barremian,  Chaine  d’Azour,  ten  from  the  Barremian  of  Oued  Aghbalou. 

Diagnosis.  Elongate  Loriolithyris , attaining  large  size;  P/A  ratio  slightly  more  than  1.  Valves  equally  convex. 
Umbo  erect.  Foramen  mesothyrid,  labiate.  Beak  ridges  rounded.  Symphytium  hidden  in  adult  stage.  Lateral 
commissure  very  strongly  arched.  Anterior  commissure  sulciplicate  with  shallow  median  sinus,  rarely  episulcate. 
Shell  folded  only  at  extreme  anterior,  marked  by  strong  concentric  growth  ridges.  Small  pedicle  collar  present. 


text-fig.  13.  Transverse  sections  through  Loriolithyris  marocensis.  Sections  5.6  and  6.0  are  enlarged  in  order  to 
show  the  detailed  structure  of  the  cardinal  process,  with  juvenile  primary  hinge  plates  (fine  stipple)  surrounded 
by  laminated  thickening  and  the  body  of  the  cardinal  process  infilled  with  punctate  skeletal  tissue  (coarse 
stipple).  Section  6.4  is  enlarged  to  show  the  primary  hinge  plates  (stippled).  Maximum  development  of  the  crural 
processes  is  seen  in  section  1 1.2.  Note  the  height  of  the  transverse  band  above  the  floor  of  the  valve  in  section 
14.4.  S. 547/1,  Gentil  Coll.,  Hauterivian,  locality  unknown.  A— scale  for  sections  5.6,  6.0,  and  6.4.  B— scale  for 

the  remaining  sections. 


530 


PALAEONTOLOGY,  VOLUME  23 


Hinge  plates  initially  concave,  becoming  rounded  L-shaped,  piped.  Cardinal  process  extends  along  the  hinge 
plates,  leaving  small  dorsal  umbonal  cavity.  Transverse  band  high-arched,  with  somewhat  pointed  crest,  high 
above  floor  of  valve. 

Remarks.  As  all  the  specimens  available  are  fully  adult  or  gerontic  little  can  be  said  about  the 
ontogeny,  except  that  biplication  of  the  anterior  commissure  and  folding  of  the  shell  appear  to 
develop  very  late.  L.  marocensis  differs  from  most  species  of  the  genus  in  the  large  size  attained  when 
adult  and  the  massive,  little-folded  form  of  the  shell;  in  those  respects  it  is  nearest  to  L.  melaitensis  but 
differs  markedly  from  that  species  in  its  internal  structures:  L.  melaitensis  is  distinguished  by  the  large 
size  of  its  crural  bases  whereas  in  L.  marocensis  the  crural  bases  are  small  and  enclosed  within  the 
piped  edge  of  the  hinge  plate  as  usual  in  Loriolithyris.  L.  marocensis  is  also  distinct  from  other  species 
of  the  genus  in  the  L-shape  developed  by  the  hinge  plates  as  seen  in  transverse  section  (text-fig.  13). 
Another  Moroccan  Lower  Cretaceous  species  which  closely  resembles  L.  marocensis  is  Cyrtothyris 
middlemissi ; the  latter  is  broader  in  relation  to  length,  and  has  a less  erect  umbo,  and  lacks  the 
loriolithyrid  boldly  arched  lateral  commissure  of  L.  marocensis,  besides  the  internal  differences. 

Distribution.  Hauterivian  and  Barremian  of  south-west  Morocco. 

Genus  boubeithyris  Cox  and  Middlemiss,  1978 
Type  species.  Terebratula  boubei  d’Archiac,  1847. 

Species  included.  T.  boubei  d’Arch.  Boubeithyris  buzzardensis  Cox  and  Middlemiss,  B.  tibourrensis  nov., 
B.  pleta  nov.  Range:  Hauterivian?,  Barremian  to  Cenomanian. 

Boubeithyris  tibourrensis  sp.  nov. 

Plate  56,  figs.  3, 4;  text-fig.  14 

Types.  Holotype,  Gentil  Collection  specimen  no.  S. 548/2/1,  from  Butte  de  Tibourr’m;  labelled  Aptian  (more 
likely  Barremian).  Dimensions:  L 20-5,  W 16-25,  T 12-5.  Paratype,  Gentil  Collection  specimen  no.  S. 552/3/1, 
Barremian,  Tibourr’m. 

Material.  Two  specimens  in  the  Gentil  Collection  from  Butte  de  Tibourr’m,  one  labelled  Aptian,  the  other 
Barremian. 

Diagnosis.  Boubeithyris  regularly  oval  as  seen  in  ventral  profile,  apart  from  short  straight  anterior  (between  the 
lateral  plicae).  Valves  equally  convex.  P/A  ratio  slightly  greater  than  1.  Umbo  suberect;  beak  ridges  moderately 
well  defined.  Foramen  mesothyrid,  marginate,  slightly  telate.  Lateral  commissure  arched.  Anterior  commissure 
sulciplicate;  lateral  plicae  close  together;  median  sinus  narrow.  Plication  reflected  by  small  folds  and  sulci  in 
extreme  anterior  part  of  brachial  valve  only.  Hinge  plates  thin,  concave,  piped  to  strongly  corniced.  Inner  socket 


EXPLANATION  OF  PLATE  56 

Figs.  1,  2.  Loriolithyris  marocensis  sp.  nov.  1 a-d,  holotype,  S. 547/2  Gentil  Coll.,  Upper  Hauterivian,  loc. 
unknown.  2 a-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  13),  S. 547/1,  Gentil  Coll.,  Upper 
Hauterivian,  loc.  unknown. 

Figs.  3,  4.  Boubeithyris  tibourrensis  sp.  nov.  3 a-d,  holotype,  S.548/2/1,  Gentil  Coll.,  Barremian  or  Aptian, 
Butte  de  Tibourr’m.  4 a-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  14),  S.522/2/1,  Gentil  Coll., 
Barremian,  Tibourr’m. 

Figs.  5,  6.  Boubeithyris  pleta  sp.  nov.  5 a-d,  holotype,  S. 553/3,  Gentil  Coll.,  Barremian,  Sidi  Bou  Rjaa.  6 a-c, 
plaster  cast  of  specimen  sectioned  (see  text-fig.  15),  S.553/1,  Gentil  Coll.,  Barremian,  Sidi  Bou  Rjaa. 

Fig.  7 a-d.  Boubeithyris  pleta  sp.  nov.  Large  typical  specimen,  S. 557/6,  Gentil  Coll.,  Barremian,  Igueni  Ouram. 

Fig.  8.  Paraboubeithyris plicae  gen.  et  sp.  nov.  8 a-d,  holotype,  S. 548/1/3,  Gentil  Coll.,  Barremian,  Vallee  Asif 
Ait  Ameur. 

All  natural  size. 


PLATE  56 


middlemiss,  Cretaceous  Terebratulidae 


532 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  14.  Transverse  sections  through  Boubeithyris  tibourrensis.  Sections  3.0  and  3.3  are  enlarged  to  show  the 
initial  shape  of  the  juvenile  hinge  plates  within  the  cardinal  process.  Cornicing  of  the  hinge  plates  is  best  seen  in 
sections  4.2-5.4.  Section  7.8  shows  the  maximum  development  of  the  crural  processes.  The  transverse  band  was 
not  preserved  in  this  specimen.  S.552/2/1,  Gentil  Coll.,  Barremian,  Tibourr’m.  A— scale  for  sections  3.0  and  3.3 
B— scale  for  the  remaining  sections. 


ridges  narrow.  Accessory  articulation  slightly  developed.  Euseptoidum  short,  confined  to  posterior  part  of  hinge 
plates,  flanked  by  lateral  ridges. 

Remarks.  This  species  closely  resembles  the  type  species  in  general  shape,  the  close-set  lateral  plicae 
being  particularly  characteristic  of  both  species.  B.  tibourrensis  differs  from  B.  boubei  in  being  more 
oval,  less  pentagonal,  in  ventral  profile  and  somewhat  more  convex  in  lateral  profile.  Like  B.  boubei,  it 
differs  from  B.  buzzardensis  in  being  narrower  and  thicker,  having  a higher  P/A  ratio  and  folding 
almost  confined  to  the  brachial  valve.  Internally  the  hinge  plates  are  more  deeply  concave  and  the 
cornice-structure  better  developed  than  in  either  B.  boubei  or  B.  buzzardensis.  A species  of 
Boubeithyris  which  occurs  in  the  Aptian  of  the  Jura  region,  so  far  undescribed,  differs  from  B. 
tibourrensis  in  being  still  more  convex  and  in  having  a lateral  commissure  still  more  strongly  arched, 
lateral  plicae  even  closer  together,  and  a longer  symphytium.  Although  only  two  specimens  are 
available,  this  species  is  important  because  it  extends  back  to  the  Barremian  the  time-range  of  the 
typical  oval  form  of  Boubeithyris,  which  can  thence  be  traced  through  the  undescribed  Aptian  species 
from  the  Jura  to  B.  boubei  itself  in  the  Albian  and  Cenomanian. 

Distribution.  Barremian  of  south-west  Morocco. 

Boubeithyris  pleta  sp.  nov. 

Plate  56,  figs.  5-7;  text-fig.  15 

Types.  Holotype,  Gentil  Collection  specimen  no.  S.553/3,  from  the  Barremian  of  Sidi  Bou  Rjaa,  Oued  Tidzi. 
Dimensions:  L 25-5,  W 23-75,  T 15.  Paratypes,  Gentil  Collection  specimens  S. 553/1  (age  and  locality  as 
holotype)  and  S. 557/6,  Barremian,  Igueni  Ouram. 


MIDDLEMISS:  CRETACEOUS  TEREB R ATU LI D AE 


533 


Material.  Twenty-one  specimens  in  the  Gentil  Collection. 

Name.  Latin  pleta,  ‘filled’,  from  the  well-filled  appearance  of  the  shell. 

Diagnosis.  Boubeithyris  almost  as  broad  as  long,  with  thickness  less  than  two-thirds  of  width.  Subcircular  in 
ventral  profile.  Valves  equally  convex.  P/A  ratio  about  1 . Umbo  short,  suberect.  Beak  ridges  rounded.  Foramen 
mesothyrid,  attrite.  Lateral  commissure  arched.  Anterior  commissure  sulciplicate;  median  sinus  low.  Shell  little 
folded.  Hinge  plates  concave,  piped  to  strongly  corniced.  Euseptoidum  short  and  weak.  Transverse  band 
moderately  high. 

Remarks.  In  external  appearance  this  species  could  be  taken  for  a sulciplicate  species  of  Sellithyris  but 
the  extremely  gentle  folding  imparts  to  the  shell  a tumid  or  ‘well-filled’  appearance  which  is 
distinctive;  also  the  ventral  profile  is  less  pentagonal  than  in  most  species  of  Sellithyris , even 
S.  deningeri  which  is  a particularly  rounded  species  of  that  genus.  It  differs  from  other  species  of 
Boubeithyris  mainly  in  being  relatively  wide  and  flat  in  comparison  with  its  length  and  in  the  wider 
spacing  of  the  plicae  of  the  anterior  commissure. 

Distribution.  Hauterivian(?)  and  Barremian  of  south-west  Morocco. 


text-fig.  15.  Transverse  sections  through  Boubeithyris  pleta.  Sections  2.8-4.8  are  enlarged  to  show  details  of  the 
structure  of  the  hinge  plates  and  of  the  cornicing.  Maximum  height  of  the  crural  processes  is  seen  in  section  6.4. 
S.553/1,  Gentil  Coll.,  Barremian,  Sidi  Bou  Rjaa.  A— scale  for  sections  2. 8-4.8.  B— scale  for  the  remaining 

sections. 


Genus  paraboubeithyris  gen.  nov. 

Type  species.  Paraboubeithyris  plicae  sp.  nov. 

Diagnosis.  Ventral  profile  rounded  pentagonal,  as  wide  as,  or  wider  than,  long.  Depressed.  P/A  ratio  slightly 
more  than  1.  Umbo  suberect  to  erect.  Beak  ridges  rounded.  Foramen  mesothyrid,  marginate,  becoming  labiate. 
Lateral  commissure  strongly  arched.  Anterior  commissure  deeply  uniplicate,  or  sulciplicate  with  very  small 
median  sinus.  Brachial  valve  has  a strong  median  fold  extending  from  the  umbonal  region  to  the  anterior; 
corresponding  to  a deep,  wide  sulcus  in  the  anterior  half  of  the  pedicle  valve.  Hinge  plates  concave,  thin,  sharply 
differentiated  from  the  inner  socket  ridges;  piped  to  strongly  corniced.  Transverse  band  high-arched. 
Euseptoidum  weak,  flanked  by  two  low  lateral  ridges. 


534 


PALAEONTOLOGY,  VOLUME  23 
Paraboubeithyris  plicae  sp.  nov. 

Plate  56,  fig.  8;  Plate  57,  figs.  1-3;  text-fig.  16 


Types.  Holotype,  Gentil  Collection  specimen  no.  S. 548/1/3,  from  the  Barremian  of  the  Vallee  Asif  Ait  Ameur. 
Dimensions:  L22,  W 22-5,  T 10.  Paratypes,  Gentil  Collection  specimens  S.546/1/1,  S. 546/1/2,  and  S.546/1/3;  age 
given  as  Upper  Hauterivian  (locality  unknown). 

Name.  Genitive  of  Latin  plica , ‘a  fold’. 

Material.  Thirty-three  specimens  in  the  Gentil  Collection,  of  which  ten  are  from  the  Barremian  of  Vallee  Asif  Ait 
Ameur  and  twelve  from  the  Barremian  of  Ida  ou  Tanan,  the  remainder  being  unlocated. 

Description.  This  species  has  a deep  and  dramatic  uniplication,  especially  in  the  more  gerontic  specimens.  Some 
of  the  smaller  specimens  have  a very  small  median  sinus,  so  that  the  anterior  commissure  is  strictly  sulciplicate, 
but  the  sinus  is  always  extremely  small  and  usually  asymmetrically  placed.  We  lack  juvenile  representatives  of  the 


text-fig.  1 6.  Transverse  sections  through  Paraboubeithyris  plicae.  Section  1 .2  shows  the  pedicle  collar  (stippled). 
Section  3.2  shows  a dorsal  umbonal  cavity.  The  corniced  hinge  plates  are  well  seen  in  sections  4.8  and 
5.2.  S.546/1/1,  Gentil  Coll.,  Hauterivian,  locality  unknown. 


EXPLANATION  OF  PLATE  57 

Figs.  1-3.  Paraboubeithyris  plicae  gen.  et  sp.  nov.  1 a-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  16), 
S.546/1/1,  Gentil  Coll.,  Upper  Hauterivian,  loc.  unknown.  2 a-d,  adult  but  uniplicate  form,  S. 546/1/2,  Gentil 
Coll.,  Upper  Hauterivian,  loc.  unknown.  3 a-d,  elongate  form  showing  incipient  biplication,  S.546/1/3, 
Gentil  Coll.,  Upper  Hauterivian,  loc.  unknown. 

Fig.  4 a-c.  Cyrtothyris  middlemissi  (Calzada),  plaster  cast  of  specimen  sectioned  (see  text-fig.  1 8),  BM  BB  76564, 
D.V.  Ager  Coll.,  Aptian,  Ait  Abaid,  Agadir. 

Figs.  5,  6.  Cyrtothyris  middlemissi  (Calzada).  5 a-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  1 7),  BM  BB 
76565,  Calzada  Coll.,  Aptian,  La  Roqueta,  Spain.  6 a-c,  BM  BB  76566,  Calzada  Coll.,  Albian,  Peracals, 
Spain. 

All  natural  size. 


PLATE  57 


middlemiss,  Cretaceous  Terebratulidae 


536 


PALAEONTOLOGY,  VOLUME  23 


species  but  specimens  in  the  Gentil  Collection  indicate  that  the  sinus  appears  late,  following  juvenile 
rectimarginate  and  uniplicate  stages,  when  the  shell  has  attained  a length  of  about  15  mm,  and  is  then  lost  again 
in  the  gerontic  stage.  Some  individuals  show  no  sign  of  biplication,  however. 

Remarks.  This  species  is  almost  certainly  the  form  that  both  Roch  and  Ambroggi  identified  as 
Terebratula  collinaria  d’Orbigny,  which  it  resembles  in  general  shape.  The  principal  differences 
between  these  two  species  are  (a)  T.  collinaria  is  always  uniplicate,  never  biplicate;  ( b ) the  cardinal 
slopes  of  T.  collinaria  tend  to  be  concave  in  dorsal  profile,  with  a sharply  produced  umbo,  those  of 
P.  plicae  are  convex,  with  an  umbo  which  does  not  protrude  beyond  the  curve  of  the  cardinal  slopes; 
(c)  T.  collinaria  has  relatively  flat  hinge  plates  with  no  trace  of  the  corniced  structure  characteristic  of 
Paraboubeithyris. 

Distribution.  Barremian  of  south-west  Morocco. 

Subfamily  rectithyridinae  Muir-Wood,  1965 
Genus  cyrtothyris  Middlemiss,  1959 

Type  species.  Terebratula  depressa  var.  cyrta  Walker,  1868. 

Species  included.  T.  depressa  var.  cyrta  Walker,  T.  depressa  var.  uniplicata  Walker,  T.  depressa  var. 
cantabridgiensis  Walker,  T.  seeleyi  Walker,  T.  dallasi  Walker,  Cyrtothyris  middlemissi  Calzada,  C.  cyrta  arminiae 
Middlemiss,  ‘ Cyrtothyris ’ maynci  Owen.  Range:  Valanginian  to  Albian. 

Cyrtothyris  middlemissi  Calzada 
Plate  57,  figs.  4-6;  text-figs.  17,  18 
* 1972  Cyrtothyris  middlemissi  Calzada,  p.  66,  fig.  1. 

Holotype.  Geological  Museum  of  the  Seminario  de  Barcelona,  specimen  no.  23.346,  from  the  Aptian  of  La 
Roqueta,  Garraf,  Barcelona. 


text-fig.  17.  Transverse  sections  through  Cyrtothyris  middlemissi.  Sections  6.0  and  6.4  are  enlarged  to  show  the 
initial  horizontal  cuneate  shape  of  the  hinge  plates.  BM  BB  76565,  Coll.  S.  Calzada,  Aptian,  La  Roqueta,  Spain. 
A— scale  for  sections  6.0  and  6.4.  B— ^ scale  for  the  remaining  sections. 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


537 


text-fig.  18.  Transverse  sections  through  Cyrtothyris  middlemissi.  Maximum  height  of  the  crural  processes  is 
seen  at  16.4.  BM  BB  76564,  Coll.  D.  V.  Ager,  Aptian,  Ait  Abaid,  Agadir,  Morocco. 


Material.  Nineteen  specimens  in  the  Gentil  Collection  (seventeen  from  the  Clansayesian  of  Sidi  Bou  Rjaa,  one 
from  the  Clansayesian  of  Imi  ou  Tanant,  one  from  the  Aptian  of  Ait  Moujjout).  Three  specimens  from  probable 
Aptian,  Ait  Abaid,  north-east  of  Agadir  (Ager  Collection).  Also  nineteen  other  specimens:  three  from  the 
Aptian  of  La  Roqueta  (Calzada  Collection);  four  from  the  Upper  Aptian,  Plan  de  Coloubret,  Taura,  Aude 
(Charriere  Collection);  six  from  the  Aptian  of  Combe  Longue,  Taura,  Aude;  two  from  the  Albian  of  Peracals, 
Lerida,  Spain  (Calzada  Collection);  four  from  the  Albian  of  Pic  du  Seigneur,  Tuchan,  Aude  (Debuyser 
Collection). 

Original  diagnosis  (after  Calzada  1972).  Large  forms  (maximum  L 53,  W 36,  T 24;  L/W  ratio  1-1-1  -6;  L/T  ratio 
1-7-21)  of  subpentagonal  to  oval  ventral  profile.  Maximum  width  and  thickness  in  middle  of  length.  Valves 
convex,  pedicle  valve  much  more  so  than  brachial  valve.  Valves  may  show  folding  (but  this  character  is  very 
variable).  Lateral  commissure  inclined  ventralwards  at  about  20°  and  arched.  Anterior  commissure  uniplicate  to 
slightly  sulciplicate.  Umbo  wide,  massive,  suberect  to  erect.  Foramen  wide,  labiate,  circular,  mesothyrid. 
Interareas  somewhat  concave;  beak  ridges  moderately  rounded.  Deltidial  plates  small  but  visible,  fused  into  a 
symphytium.  Growth  lines  visible.  Hinge  plates  concave,  somewhat  clubbed,  becoming  anteriorly  persistently 
virgate  or  even  V-shaped.  Angle  between  the  crural  bases  and  the  crural  rami  70°-100°.  Loop  strongly  recurved 
in  a posterior  direction  so  that  no  one  serial  section  includes  the  whole  of  the  arch  of  the  transverse  band. 

Remarks.  Specimens  from  Morocco  and  from  the  Albian  of  north-east  Spain  exceed  Calzada’s  stated 
maximum  width  (up  to  43  mm);  nevertheless  all  specimens  available  fall  into  the  range  of  L/W  ratios 
given  in  his  diagnosis.  On  the  other  hand  specimens  from  both  areas,  and  including  the  type  locality, 
fall  outside  the  range  of  L/T  ratios  given  (extremes  are  specimen  MDA  2/1 , from  Morocco,  1 -57  and 
CaP2,  from  the  Albian  of  Peracals,  2-12).  Calzada  understates  the  plication  of  the  anterior 
commissure,  which  is  normally  gently  sulciplicate  in  the  adult  stage.  The  foramen  should  be  described 


538 


PALAEONTOLOGY,  VOLUME  23 


as  strongly  marginate,  labiate  in  the  adult  stage.  The  wide  triangular  shape  of  the  loop  and  the  strong 
recurvature  of  the  transverse  band  are  generic  features  in  Cyrtothyris  (Middlemiss  1976). 

Distribution.  Aptian  of  Aude  and  north-eastern  Spain;  Aptian  (including  Clansayesian)  of  south- 
western Morocco;  Albian  of  Aude  and  north-eastern  Spain. 

Subfamily  uncertain 
Genus  kutchithyris  Buckman,  1918 

Type  species.  Terebratula  acutiplicata  Kitchin,  1900. 

Original  definition  (Buckman  1918).  ‘Permesothyrid  (beak  stout,  broad,  quite  short,  thickened  with  callus, 
obliquely  truncate,  foramen  large,  circular,  attrite,  close  to  umbo,  symphytium  very  short);  morphogeny, 
biconvex  to  strongly  sulciplicate;  muscle-tracks  obliterated  posteriorly,  not  reaching  far  down  valves,  rather 
sharply  divergent,  starting  not  from  the  umbo  but  from  about  midway  of  the  posterior  half  of  the  shell,  showing 
little  more  than  scars;  dorsal  septum  feeble— ovarian  areas  large,  mammillate  on  cast.  The  muscle  scars 
posteriorly  obliterated  and  diverging  from  a point  well  removed  from  the  umbo,  the  short  beak  with  little 
exposure  of  symphytium:  these  characters  at  once  distinguish  the  genus.’ 

Diagnosis.  Umbo  suberect  to  incurved.  Foramen  mesothyrid  to  epithyrid;  may  be  slightly  labiate.  Development 
of  anterior  commissure  uniplicate  to  sulciplicate,  more  rarely  to  episulcate.  Hinge  plates  wide,  concave, 
flattening  anteriorly,  very  little  differentiated  from  the  laterally  deflected  inner  socket  ridges.  Crural  bases  low 
where  attached  to  hinge  plates,  rapidly  elongating  anteriorly  and  passing  into  high,  thin,  slightly  flanged  crural 


text-fig.  19.  Transverse  sections  through  Kutchithyris  acutiplicata  (type  species  of  the  genus).  Sections  4.8-6.0 
are  enlarged  in  order  to  show  details  of  the  structure  of  the  cardinal  process.  The  crural  bases  first  appear  at  6.0. 
The  transverse  band  at  13.6  is  broken  and  partially  displaced.  BM  52420,  Putchum  Group  (Upper  Jurassic), 
Jumara,  Kutch,  India.  A— scale  for  sections  4.8-6.0.  B— scale  for  the  remaining  sections. 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


539 


5 MM 


text-fig.  20.  Transverse  sections  through  Kutchithyris  subsella.  Sections  4.7-5.9  are 
enlarged  to  show  the  initial  shape  of  the  hinge  plates  at  4.7  and  5.1,  the  primary  hinge 
plates  (stippled)  at  5.5,  and  the  first  appearance  of  the  crural  bases  at  5.9.  Maximum 
height  of  the  crural  processes  is  seen  at  7.9.  The  transverse  band  was  not  preserved  in 
this  specimen.  BM  BB  76555,  Kimeridgian,  Le  Havre,  France.  A — scale  for  sections 
4.7-5.9.  B — scale  for  the  remaining  sections. 


processes.  Hinge  plates  and  crural  processes  usually  clubbed.  Descending  lamellae  thin.  Transverse  band  high- 
arched,  ogival.  Euseptoidum  present  but  usually  weak;  may  be  bounded  by  two  low  euseptoidum-like  ridges 
bounding  the  adductor  impressions. 

Remarks.  The  species  here  ascribed  to  this  genus  differ  one  from  another  considerably  in  external 
proportions,  from  the  highly  convex  globular  form  of  Kutchithyris  brivesi,  through  the  pentagonal 
ventral  profile  of  K.  acutiplicata  and  K.  subsella  to  the  elongate  form  of  K.  kennedyi.  They  are  linked, 
however,  by  close  similarity  in  the  internal  characters,  especially  those  of  the  hinge  plates,  inner 
socket  ridges,  and  crural  bases.  Buckman  erected  the  genus  Kutchithyris  mainly  to  accommodate  six 
species  from  the  Bathonian  and  Callovian  of  India  previously  established  by  Kitchin  but  he  also 
included  two  European  species  of  Deslongchamps  and  two  newly  established  species  of  his  own  from 
the  English  Great  Oolite  (Bathonian)  of  Bradford-on-Avon,  K.  fulva  and  K.  egregia. 

I here  refer  to  Kutchithyris  the  species  T.  subsella  Leymerie,  a familiar  Upper  Jurassic  species  in 
Europe,  which  has  been  previously  referred  to  Sellithyris  by  Barczyk  (1969).  I exclude  it  from 
Sellithyris  mainly  because  of  the  lack  of  differentiation  between  hinge  plates  and  inner  socket  ridges, 
the  detailed  form  of  the  hinge  plates  (as  seen  in  transverse  section  they  are  like  hockey  sticks),  and  the 
form  of  the  crural  processes;  these  are  features  which  it  shares  with  other  species  of  Kutchithyris. 
K.  subsella  survived  into  the  Lower  Cretaceous  and  occurs  in  the  Upper  Valanginian  of  La  Querola 


540 


PALAEONTOLOGY,  VOLUME  23 


. 70 


text-fig.  21 . Transverse  sections  through  Kutchithyris  subsella.  Sections  4.2  and  4.6  are  enlarged  to  show  detail 
of  the  primary  hinge  plates.  The  crural  bases  are  first  seen  at  5.0.  The  crural  processes  are  at  their  maximum 
height  at  9.0.  The  transverse  band  was  not  preserved  in  this  specimen.  BM  BB  76558,  Coll.  M.  Durand  Delga, 
Niveau  14A,  Valanginian,  La  Querola,  Spain.  A — scale  for  sections  4.2  and  4.6.  B — scale  for  the  remaining 

sections. 


EXPLANATION  OF  PLATE  58 

Figs.  1 -6.  Kutchithyris  kennedyi  sp.  nov.  1 a-d,  holotype,  BM  BB  76556,  Y.  Champetier  Coll.,  Hauterivian  or 
Barremian,  Oliva,  Valencia,  Spain.  2 a-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  23),  BM  BB  76557, 
Y.  Champetier  Coll.,  Hauterivian  or  Barremian,  Oliva,  Valencia,  Spain.  3 a-d,  BM  BB  76559,  Durand  Delga 
Coll.,  Valanginian,  La  Querola,  Alicante,  Spain.  4 a-c,  typical  specimen,  BM  BB  76562,  W.  J.  Kennedy  Coll., 
Lower  Barremian,  Les  Moulins,  Mont  Chauve,  Nice,  France.  5 a-c,  large  adult  specimen,  plaster  cast  of 
specimen  sectioned  (see  text-fig.  24),  BM  BB  76561,  Y.  Rangheard  Coll.,  ?Hauterivian,  Punta  Torreta,  Ibiza. 
6 a-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  22),  S. 552/1/1,  Gentil  Coll.,  Hauterivian,  Ifrech-Oued- 
Igouzoulen. 

Figs.  7-9.  Kutchithyris  brivesi  (Roch).  la-c,  plaster  cast  of  specimen  sectioned  (see  text-fig.  26),  S. 549/2,  Gentil 
Coll.,  Hauterivian,  Ifrech-Oued-Igouzoulen.  8 a-d,  uniplicate  specimen,  S.549/3,  Gentil  Coll.,  Hauterivian, 
Ifrech-Oued-Igouzoulen.  9 a-d,  gerontic  episulcate  specimen,  S. 549/4,  Gentil  Coll.,  Hauterivian,  Ifrech- 
Oued-Igouzoulen. 

All  natural  size. 


PLATE  58 


middlemiss,  Cretaceous  Terebratulidae 


542 


PALAEONTOLOGY,  VOLUME  23 


north  of  Alcoy,  Alicante,  Spain  (Durand  Delga  Collection).  The  other  Cretaceous  species  of  the 
genus,  which  are  described  here,  are  new. 

Species  included.  Bathonian:  T.  hypsogonia  Kitchin,  T.  acutiplicata  Kitchin,  T.  propinqua  Kitchin,  T. 
circumdata  Deslongchamps,  IK.  fulva  Buckman,  IK.  egregia  Buckman.  Callovian:  T.  aurata 
Kitchin,  T.  jooraensis  Kitchin,  IT.  longicarinata  Kitchin,  T.  subcanaliculata  Deslongchamps. 
Oxfordian  to  Valanginian:  T.  subsella  Leymerie.  Valanginian  to  Barremian:  K.  kennedyi  nov., 
K.  brivesi  (Roch). 

Range  of  the  genus.  Bathonian  to  Barremian. 

Kutchithyris  kennedyi  sp.  nov. 

Plate  58,  figs.  1-6;  text-figs.  22-24 

Types.  Holotype,  BM  BB  76556,  from  Oliva,  Valencia,  Spain  (Champetier  Collection).  The  horizon  is  dubious 
but  is  probably  Hauterivian  or  Barremian.  Dimensions:  L 30,  W 20,  T 18-5.  Paratypes.  BM  BB  76557,  Oliva, 
Valencia,  Spain;  BM  BB  76559,  Upper  Valanginian,  La  Querola,  Alicante,  Spain;  BM  BB  76561,  PuntaTorreta, 
Ibiza;  BM  BB  76562  and  76563,  Lower  Barremian,  Mont  Chauve,  Alpes  Maritimes,  France;  Gentil  Collection 
S. 552/1/1,  Hauterivian,  Ifrech  Oued  Igouzoulen,  Morocco. 

Material.  Three  specimens  from  Oliva,  Valencia,  Spain  (Champetier  Collection,  horizon  uncertain).  Five 
specimens  from  niveau  14A  at  La  Querola,  north  of  Alcoy,  Alicante,  Spain  (Busnardo  and  Durand  Delga  1960) 


b 

text-fig.  22.  Transverse  sections  through  Kutchithyris  kennedyi.  Section  3.6  is  enlarged  to  show  the  juvenile 
primary  hinge  plates  within  the  cardinal  process.  The  crural  bases  are  first  seen  at  4.4  and  maximum  development 
of  the  crural  processes  at  7.2.  Sections  3. 6-6.0  S.552/1/1;  sections  6.8-10.0  S. 552/1/2.  Both  specimens  Gentil 
Coll.,  Hauterivian,  Ifrech-Oued-Igouzoulen.  A— scale  for  section  3.6.  B— scale  for  the  remaining  sections. 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


543 


(Durand  Delga  Collection,  probably  Valanginian).  Two  specimens  from  the  Lower  Barremian  of  a stream 
section  800  m north  of  Les  Moulins,  east  of  Mont  Chauve,  north  of  Nice,  Alpes  Maritimes  (Kennedy 
Collection).  One  specimen  from  Ecru,  Morocco  (Whitaker  Collection).  One  specimen  from  Punta  Torreta,  Ibiza 
(Rangheard  Collection,  probably  Hauterivian).  Four  specimens  in  the  Gentil  Collection  (three  from  the 
Hauterivian  of  Ifrech  Oued  Igouzoulen,  one  from  the  Barremian  of  Asif  Ait  Ameur). 

Name.  Named  after  Dr.  W.  J.  Kennedy,  who  supplied  some  of  the  specimens. 

Diagnosis.  Kutchithyris  of  elongate  oval  ventral  profile  (width  about  0-7  length);  thickness  more  than  half  length. 
P/A  ratio  1 -3—1-6.  Umbo  suberect  to  erect  in  adults.  Symphytium  very  short  or  invisible.  Foramen  mesothyrid, 
labiate.  Beak  ridges  rounded.  Anterior  commissure  sulciplicate  to  episulcate.  Folding  of  the  shell,  corresponding 
to  the  plicae  and  sinuses  of  the  commissure,  weak  and  confined  to  the  anterior  third  of  the  shell  except  in  gerontic 
stage. 

Description.  Because  of  the  few  specimens  available  little  can  be  said  about  the  ontogeny  of  this  species  except 
that  the  width/length  ratio  appears  to  be  isometric  and  to  remain  constant  during  growth  at  a little  less  than  0-7, 
whereas  the  thickness/length  ratio  is  allometric. 


text-fig.  23.  Transverse  sections  through  Kutchithyris  kennedyi.  Sections  3. 2-4.0  are 
enlarged  in  order  to  show  the  juvenile  hinge  plates  within  the  cardinal  process  (at  3.2)  and 
the  crural  bases  (at  3.6  and  4.0).  Maximum  height  of  the  crural  processes  is  seen  at  8.0.  The 
transverse  band  was  not  preserved  in  this  specimen.  BM  BB  76557,  Coll.  Y.  Champetier, 
Oliva,  Spain.  A — scale  for  sections  3.2-4.0.  B— scale  for  the  remaining  sections. 


544 


PALAEONTOLOGY,  VOLUME  23 


Remarks.  This  species  is  easily  distinguished  from  other  members  of  Kutchithyris  by  its  elongate 
form.  The  species  with  which  it  is  most  likely  to  be  confused  is  Loriolithyris  valdensis.  K.  kennedyi  is 
thicker  in  relation  to  its  length  than  L.  valdensis,  because  the  differential  growth  ratio  of  this  character 
is  slightly  bigger,  giving  the  allometric  distribution  a slightly  steeper  slope  (fig.  8).  In  addition,  the 
brachial  valve  of  K.  kennedyi  is  slightly  concave  in  anterior  third,  that  of  L.  valdensis  uniformly 
convex  in  lateral  view.  Internally  the  characters  of  the  hinge  plates,  inner  socket  ridges,  and  crural 
bases  are  all  quite  different  in  the  two  species. 

Distribution.  ?Valanginian  of  south-east  Spain;  Hauterivian  and  Barremian  of  south-west  Morocco; 
?Hauterivian  of  Ibiza;  Lower  Barremian  of  south-east  France. 


text-fig.  24.  Transverse  sections  through  a large,  adult  specimen  of  Kutchithyris  kennedyi.  Sections  5.2  and  5.6 
are  enlarged  to  show  the  juvenile  hinge  plates  (at  5.2)  and  the  primary  hinge  plates  (stippled  at  5.6).  The  crural 
bases  are  already  visible  at  5.6.  BM  BB  76561,  Coll.  Y.  Rangheard,  Punta  Torreta,  Ibiza.  A — scale  for  sections 
5.2  and  5.6.  B — scale  for  the  remaining  sections. 


Kutchithyris  brivesi  (Roch) 

Plate  59,  figs.  1,  2;  text-figs.  25,  26 

v*  1930  Terebratula  brivesi  Roch,  p.  259,  pi.  22,  figs.  12-13. 
vl951  Terebratula  brivesi  Roch;  Gigout,  p.  361,  pi.  9,  figs.  27-34. 

Lectotype.  Roch  figured  two  specimens  but  there  is  confusion  in  the  numbering  of  the  figures;  figs.  12a  and  13 b j[ 

represent  one  specimen,  figs.  1 2b  and  1 3 a the  other.  The  specimen  represented  by  figs.  1 2a  and  1 3 b is  here  chosen 
as  lectotype.  It  is  in  the  collection  of  the  Service  de  la  Carte  Geologique  du  Maroc  at  Rabat,  bearing  the  number 
Ci  55,  and  is  from  the  Valanginian  of  Zauouia  Embarek  des  Ida  ou  Troumma.  The  label  describes  it  as  ‘Coll.  E.  ; 
Roch’  but  Roch  in  his  caption  gives  it  as  ‘Brives  Coll.’. 

Paratypes.  The  specimen  figured  by  Roch  as  figs.  12 b and  13a  (at  Rabat,  bearing  the  same  number  as  the  lecto- 
type and  from  the  same  horizon  and  locality).  A specimen  in  the  Roch  Collection  at  Rabat  bearing  number  P 62 
and  coming  from  the  Berriasian  of  Dar  Caid  Tigzirin.  Six  specimens  in  the  Roch  Collection  at  Rabat  bearing  the 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


545 


number  P 50  and  coming  from  the  Valanginian  of  Oued  Igoulouzen.  The  following  specimens  in  the  Gentil  Col- 
lection: S. 549/1,  S. 549/2,  S. 549/3,  S. 549/4,  S. 549/5,  S. 559/1,  all  labelled  Hauterivian,  Ifrech-Oued-Igoulouzen. 
The  two  specimens  figured  by  Gigout  (both  numbered  720  in  the  Gigout  Collection,  Universite  Mohamed  V, 
Rabat). 

Material.  Nine  specimens  from  the  Roch  Collection  (detailed  above).  Forty-eight  specimens  from  the  Gentil 
Collection  (forty-five  labelled  Hauterivian  of  Ifrech-Oued-Igoulouzen;  three  labelled  Barremian,  Chaine 
d’Azour). 

Diagnosis.  Kutchithyris  highly  obese  in  lateral  profile,  oval  in  ventral  profile.  P/A  ratio  slightly  more  than  1 . 
Brachial  valve  more  convex  than  pedicle  valve.  Umbo  erect  to  incurved.  Symphytium  very  short  to  invisible. 
Foramen  mesothyrid,  labiate  in  older  individuals.  Beak  ridges  rounded.  Lateral  commissure  arched.  Anterior 
commissure  rectimarginate  to  sulciplicate  or  episulcate.  Shell  tumid  and  little  folded,  or  not  folded. 
Euseptoidum  well  developed  in  the  region  of  the  hinge  plates  and  flanked  by  two  lateral  ridges. 


25r 


text-fig.  25.  Scatter  diagrams  of  the  relationships  of  thickness  to  length  and  thickness  to  width  in  Kutchithyris 

brivesi  (Gentil  Coll.). 


Description.  The  growth  of  this  species  is  accompanied  by  rapid  increase  in  the  thickness/length  ratio.  In  the  most 
adult  individuals  thickness  can  exceed  width.  The  smallest  specimens  available  (L  1 8-5)  are  either  rectimarginate 
or  gently  uniplicate  but  the  later  development  of  the  commissure  is  the  most  variable  character  of  the  species. 
Some  specimens  of  29  mm  in  length  are  clearly  and  deeply  uniplicate,  while  other  specimens  of  similar  size  are 
sulciplicate  or,  rarely,  episulcate.  In  other  specimens  again  a clearly  episulcate  commissure  is  developed  at  a shell 
length  of  as  little  as  19-5  mm. 

Remarks.  This  species  is  distinguishable  at  once  from  other  species  of  Kutchithyris  and  from  all  the 
other  species  considered  here  by  its  globular  form  and  the  tumid  appearance  of  both  valves. 
Internally  it  differs  from  other  species  of  Kutchithyris  in  having  a well-developed,  although  short, 
euseptoidum.  Both  Roch  and  Gigout  underestimate  the  plication  which  the  anterior  commissure 
may  show  in  this  species.  Roch  states:  ‘La  commissure  frontale  est  pratiquement  droite,  sauf  deux 
petits  plis  a peine  marques.’  According  to  Gigout:  ‘Commissure  frontale  droite  ou  tres  legerement 
convexe  vers  la  petite  valve.’  The  larger  specimens  (L  25-5)  in  Roch’s  own  collection,  however,  are 
strongly  uniplicate.  The  form  of  the  anterior  commissure  of  the  larger  specimens  in  the  Gentil 
Collection  is  very  variable,  suggesting  that  Roch  and  Gigout  may  have  seen  only  small,  relatively 
juvenile  specimens  such  as  the  lectotype.  Roch,  Gigout,  and  Ambroggi  all  give  the  main  occurrence  of 
this  species  as  of  Valanginian  age,  Roch  and  Ambroggi  recording  some  also  from  the  Berriasian, 
whereas  the  great  majority  of  the  Gentil  Collection  specimens  are  labelled  Hauterivian,  with  a few 
labelled  Barremian.  It  is  possible  that  strong  sulciplication  or  episulcation  was  developed  in  this 
species  only  after  the  Valanginian.  The  unity  of  the  species  is  demonstrated  by  the  remaining 


PLATE  59 


middlemiss,  Cretaceous  Terebratulidae 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


547 


text-fig.  26.  Transverse  sections  through  Kutchithyris  brivesi.  Section  4.4  is  enlarged  to  show 
detail  of  the  structure  of  the  cardinal  process.  The  crural  bases  are  first  seen  at  4.8.  The  crural 
processes  are  at  their  maximum  height  at  7.2.  S.549/2,  Gentil  Coll.,  Hauterivian,  Ifrech-Oued- 
Igouzoulen.  A— scale  for  section  4.4.  B— scale  for  the  remaining  sections. 


EXPLANATION  OF  PLATE  59 

Figs.  1,  2.  Kutchithyris  brivesi  (Roch).  1 a-d,  juvenile  but  incipiently  biplicate  specimen,  S. 549/5,  Gentil  Coll., 
Hauterivian,  Ifrech-Oued-Igouzoulen.  2 a-d,  adult  but  uniplicate  specimen,  S.559/1,  Gentil  Coll., 
Hauterivian,  Ifrech-Oued-Igouzoulen. 

Figs.  3-7.  Juralina  ecruensis  sp.  nov.  3 a-d,  holotype,  BM  BB  76547,  Whitaker  Coll.  4 a-d,  typical  uniplicate 
form,  BM  BB  76548.  5 a-d,  plaster  cast  of  specimen  sectioned  (see  text-fig.  28),  BM  BB  76550.  6a-d,  juvenile 
specimen,  BM  BB  76551.  la-d,  elongate  adult  form,  BM  BB  76553. 

All  natural  size. 


548 


PALAEONTOLOGY,  VOLUME  23 


characters  both  external  and  internal.  A specimen  from  Roch’s  collection  (from  the  Valanginian  of 
Oued  Igouzoulen)  was  serially  sectioned  and  differed  slightly  from  the  Gentil  specimen  shown  in  text- 
fig.  26  in  having  hinge  plates  less  concave  in  their  earlier  stages,  a less  developed  euseptoidum,  and  in 
lacking  any  clubbed  thickening  of  the  hinge  plates  and  crural  processes.  These  are  signs  of 
immaturity,  confirming  that  the  specimens  described  by  Roch  were  comparatively  juvenile. 

Distribution.  Berriasian  to  Barremian  of  south-west  Morocco. 

Genus  juralina  Kyansep,  1961 
Type  species.  Juralina  procerus  Kyansep. 

Original  diagnosis  (from  Kyansep  1961).  ‘Shell  plano-convex  to  biconvex.  Anterior  commissure  rectimarginate 
to  uniplicate.  Umbo  massive,  straight  to  erect.  Deltidium  high.  Socket  ridges  high.  Cardinal  process  well 
developed  and  separated  from  the  floor  of  the  dorsal  valve.  Hinge  plates  divided,  very  narrow,  in  close  proximity 
to  the  socket  ridges.  Crural  bases  given  off  ventrally  from  the  hinge  plates.  Crura  narrow,  with  well-developed, 
sharp-pointed  crural  processes.  Loop  about  one-third  of  the  length  of  the  dorsal  valve,  triangular,  with  arched 
transverse  band.  Pedicle  collar  shaped  like  a ring  valve.  Hinge  teeth  massive,  without  denticulae.  Adductor 
muscle  impressions  oval  triangular,  narrowing  to  fine  lines  posteriorly.  Euseptoidum  small.  Shell  smooth, 
punctate.’ 

Emended  diagnosis.  Shell  plano-convex  to  biconvex,  depressed  (thickness/length  ratio  low),  subcircular  in 
ventral  profile.  Umbo  straight  to  erect.  Foramen  mesothyrid,  slightly  labiate.  Lateral  commissure  oblique  to 
arched;  anterior  commissure  rectimarginate  to  squarely  uniplicate  or  slightly  sulciplicate.  Cardinal  process  well 
developed.  Hinge  plates  rectangularly  virgate  (that  is,  L-shaped  in  cross-section  with  an  inner  lamina  at  right 
angles  to  the  outer  lamina);  clubbed.  Crural  bases  given  off  from  the  anterior  ventral  extremities  of  the  hinge 
plates.  Crural  processes  high,  sharp-pointed,  incurved  at  their  extremities.  Loop  broad;  transverse  band  high- 
arched,  arcuate  to  trapezoidal. 

Remarks.  Kyansep  considered  that  his  new  genus  strongly  resembled  Lobothyris  Buckman  but 
Juralina  differed  in  having  very  narrow  hinge  plates,  high  socket  ridges,  and  well-developed  crural 
processes,  in  lacking  a septum  to  its  pedicle  collar,  and  in  the  elliptical  shape  of  its  ventral  umbonal 
cavity.  Boullier  (1976)  has,  however,  pointed  out  several  additional  differences.  Kyansep  also 
correctly  pointed  to  a marked  external  resemblance,  but  equally  marked  internal  differences,  between 
Juralina  and  Rectithyris  Sahni.  In  addition  to  his  new  species,  Kyansep  included  in  Juralina  several 
species  from  the  Jurassic  of  Europe:  Terebratula  rauraca  Rollier,  T.  repelliniana  D’Orbigny, 
T.  censoriensis  Rollier,  T.  bullingdonensis  Rollier,  T.  cotteaui  Douville,  and  T.  moravica  Glocker.  Of 
these,  T.  moravica  was  referred  to  a new  genus  Weberithyris  by  Smirnova  (1969).  In  her  discussion  of 
the  genus  Boullier  (1976)  rejects  affinities  with  Lobothyris , Weberithyris , Tropeothyris  Smirnova,  and 
Postepithyris  Makridin  but  finds  considerable  resemblance  to  Cyrtothyris  Middlemiss.  Boullier 
added  three  more  previously  established  species — T.  bauhini,  T.  valfinensis,  and  T.  subformosa. 

Barczyk  (1969)  added  the  following  species  from  Upper  Jurassic  rocks  of  the  Holy  Cross 
Mountains  of  Poland  to  Juralina:  T.  insignis  insignis  Schiibler,  1 830,  T.  insignis  maltonensis  Oppel, 
1858,  T.  immanis  immanis  Zejszner,  1856,  T.  immanis  speciosa  Schlosser,  1882.  Of  these,  Boullier 
(1976)  has  since  referred  T.  insignis  var.  maltonensis  Oppel  to  the  genus  Galliennithyris  as 
G.  maltonensis. 

I introduced  the  terms  inner  and  outer  lamina  in  1959  and  defined  them  as  follows:  ‘A  virgate  hinge 
plate  is  divisible  into  two  parts,  the  outer  lamina  from  the  socket  ridge  to  the  virgation  and  the  inner 
lamina  on  the  inner  (median)  side  of  the  virgation.’  The  accompanying  figure  (Middlemiss  1959,  text- 
fig.  1 j),  however,  showed  cuneate  hinge  plates  with  large  crural  bases.  Because  of  this  confusion  I later 
withdrew  the  terms  inner  lamina  and  outer  lamina  (Dieni  et  al.  1975;  Middlemiss  1976).  Now  that 
more  is  known  about  the  detailed  structure  of  terebratulid  hinge  plates  (Cox  and  Middlemiss  1978) 
the  terms  are  seen  to  be  useful  in  their  original  sense  and  I use  them  here. 

Species  included.  J.  procerus  Kyansep,  ITerebratula  rauraca  Rollier,  IT.  repelliniana  d’Orbigny, 
J.  graciosa  Kyansep,  IT.  censoriensis  Rollier,  T.  bullingdonensis  Rollier,  J.  naklivkini  Kyansep, 


MIDDLEMISS:  CRETACEOUS  TEREBR ATULIDAE 


549 


T.  cotteaui  Douville,  J.  babugani  Kyansep,  J.  earns  Kyansep,  T.  bauhini  Etallon,  T.  valfinensis  de 
Loriol,  T.  subformosa  Rollier,  J.  ecruensis  nov. 

Range  of  the  genus.  Middle  Oxfordian  to  Barremian. 

Juralina  ecruensis  sp.  nov. 

Plate  59,  figs.  3-7;  text-figs.  27,  28 

Types.  Holotype,  BM  BB  76547,  Whitaker  Collection.  Dimensions:  L 34-5,  W 27,  T 18.  Paratypes:  Whitaker 
Collection  specimens  BM  B 17273,  B 17277,  BB  76548,  BB  76550,  BB  76551,  BB  76553. 

Material.  Forty-six  specimens  in  the  Whitaker  Collection.  Forty-two  specimens  in  the  Gentil  Collection  (twenty- 
five  from  the  Berriasian  or  Yalanginian  of  Tinirt  Ait  Ameur,  two  from  the  Hauterivian  of  an  unnamed  locality, 
three  from  the  Barremian  of  Igueni  Ouram,  twelve  from  the  probable  Barremian  of  Oued  Aghbalou). 

Diagnosis.  Juralina  of  subcircular  to  oval  ventral  profile;  maximum  width  about  the  mid-line;  valves  equally 
convex.  Umbo  erect.  Foramen  mesothyrid,  marginate,  becoming  labiate.  Beak  ridges  rounded.  Symphytium 
short,  hidden  in  adult  stage.  Shell  smooth,  with  faint  growth  lines.  Lateral  commissure  oblique  to  arched. 
Anterior  commissure  rectimarginate  to  squarely  uniplicate  or  slightly  sulciplicate.  Euseptoidum  absent  or 
negligible.  Transverse  band  high-arched,  rounded. 

Description.  Juvenile  specimens  resemble  the  adults  except  in  being  rectimarginate.  At  a length  of  about  22  mm 
the  characteristic  adult  uniplicate  commissure  begins  to  develop.  In  adults  over  about  30  mm  in  length  the 


30  - 
25 


10 


“5  10  15  20  25  30  35  40  45 

Length 


i 10~ 15" 


20  25  30 

Length 


35  40  45 


25 


20 


5 10  15  20  25  30  35 

Width 


text-fig.  27.  Scatter  diagrams  of  the  relationships  of  simple  dimensions  in  Juralina  ecruensis  (Whitaker  Coll.). 


550 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  28.  Transverse  sections  through  Juralina  ecruensis.  The  first  two  sections  (upper  left)  are  enlarged  in 
order  to  show  detail  of  the  structure  of  the  cardinal  process.  Sections  3. 6-5. 2 are  enlarged  to  show  the  form  of  the 
hinge  plates  and  of  the  crural  bases.  Maximum  height  of  the  crural  processes  is  seen  at  8.0.  BM  BB  76550  except 
that  4. Ox  is  from  BM  B 1 7273  and  3.2  from  BM  B 1 7277  as  these  showed  better  the  details  of  the  cardinal  process 
(Whitaker  Coll.).  A — scale  for  sections  3.2  and  4.0x.  B — scale  for  sections  3. 6-5. 2.  C — scale  for  the  remaining 

sections. 


uniplica  may  be  angular,  the  commissure  horizontal  in  the  centre;  or  it  may  develop  a gentle  sinus  in  the  centre, 
giving  a slightly  sulciplicate  stage.  The  other  main  gerontic  development  is  that  the  foramen  becomes  labiate  in 
specimens  over  about  30  mm  in  length.  Text-fig.  27  shows  that  there  are  a few  long,  narrow  variants  and  others 
that  are  exceptionally  thick. 

Remarks.  This  species  is  referred  to  Juralina  because  of  ( a)  its  external  appearance,  the  distinctive 
elements  of  which  are  the  biconvex  but  moderately  depressed  form  and  the  erect  umbo;  ( b ) the 
internal  characters,  especially  the  L-shaped  form  of  the  hinge  plates  in  transverse  section,  with  the 
crural  bases  developed  in  the  extreme  ventral  tips  of  the  inner  laminae  in  the  anterior  parts  of  the 
hinge  plates  only.  All  these  characters  appear  closely  comparable  to  those  described  and  figured  by 
Kyansep  (1961),  Barczyk  (1969),  and  Boullier  (1976). 

Distribution.  Valanginian  to  Barremian  of  south-west  Morocco. 


EXPLANATION  OF  PLATE  60 

Fig.  1.  Loriolithyris  melaitensis  sp.  nov.  Section  4.8  of  text-fig.  12  photographed  to  show  the  shape  of  the 
juvenile  hinge  plates  and  the  distinction  between  punctate  and  inpunctate  skeletal  tissue  within  the  cardinal 
process. 

Fig.  2.  Loriolithyris  marocensis  sp.  nov.  Part  of  section  6.0  of  text-fig.  13  photographed  to  show  the  primary 
piped  hinge  plate  with  its  secondary  clubbed  thickening  and  the  structure  of  the  cardinal  process. 

Fig.  3.  Boubeithyris  pleta  sp.  nov.  Part  of  section  4.0  of  text-fig.  1 5 enlarged  to  show  the  detailed  structure  of 
the  junction  between  hinge  plate  and  inner  socket  ridge. 

Fig.  4.  Kutchithyris  acutiplicata  (Kitchin).  Part  of  section  6.0  of  text-fig.  19  enlarged  to  show  the  primary  hinge 
plate  with  its  clubbed  thickening  and  the  incipient  crural  base,  all  enclosed  within  the  cardinal  process. 

Linear  scale  = 2 mm. 


PLATE  60 


middlemiss,  Cretaceous  Terebratulidae 


552 


PALAEONTOLOGY,  VOLUME  23 


TEREBRATULID  SPECIES  OF  MORE  DOUBTFUL  OCCURRENCE  IN  THE 
LOWER  CRETACEOUS  OF  SOUTH-WEST  MOROCCO 

Terebratula  sueuri  Pictet  is  recorded  by  Gigout  from  the  Valanginian  and  Hauterivian  at  Safi  and  by 
both  Roch  and  Ambroggi  from  the  Barremian.  T.  sueuri  is  a Jura  species  which  is  also  found  rarely  in 
the  Hauterivian  of  the  Lower  Saxon  Basin.  Three  specimens  in  the  Gentil  Collection,  S. 544/1  (from 
Safi),  S. 547/2/1,  and  S. 547/2/2  (both  from  the  Barremian  of  Ait  el  Faci)  have  a close  external 
resemblance  to  this  species  and  probably  represent  the  form  to  which  the  name  was  applied  by 
previous  authors.  Serial  sectioning  proved  these  to  be  an  undescribed  species  of  terebratellidine, 
which  also  occurs  in  the  Jura  region  (Collections  of  the  Institut  de  Geologie,  Neuchatel).  Gigout’s 
figured  specimen  (Gigout  1951,  pi.  9,  figs.  19-22)  has  a well-developed  dorsal  median  septum  and  is 
almost  certainly  the  same  terebratellidine  species.  The  occurrence  of  these  two  externally  similar  but 
quite  unrelated  species  together  in  the  Jura  region  is  a good  example  of  homochronous 
homoeomorphy. 

Terebratula  collinaria  d’Orbigny  is  recorded  by  both  Roch  and  Ambroggi  from  the  Hauterivian 
and  Barremian  and  by  Roch  from  the  Valanginian  also.  The  records  probably  refer  to  Para- 
boubeithyris  plicae,  although  the  Gentil  Collection  contains  specimens  of  this  species  only  from  the 
Barremian. 

Tropeothyris  salevensis  (de  Loriol).  This  is  recorded  by  Gigout  from  the  Valanginian  of  the 
environs  of  Safi  and  by  Ambroggi  from  the  Barremian  of  his  area.  On  first  viewing  the  collections  I 
referred  to  T.  salevensis  the  specimens  which  I have  here  named  Loriolithyris  melaitensis;  Gigout’s 
figured  specimen  (Gigout  1951,  pi.  9,  figs.  15-18)  is  apparently  similar  to  these  externally  except  that 
it  is  a gerontic  specimen.  The  records  probably  refer  to  L.  melaitensis. 

Moutonithyris  moutoniana  (d’Orbigny)  is  recorded  by  Roch  from  the  Barremian  and  by  Gigout 
from  the  ‘Neocomian’  and  Aptian  of  Safi  and  Sidi  Bou  Zid.  Although  Gigout  gives  in  synonymy 
Pictet’s  (1872)  figure  of  the  species,  not  d’Orbigny’s  original,  his  own  figured  specimen  looks 
reasonably  convincing  (Gigout  1951,  pi.  9,  figs.  23-26).  In  the  Gentil  Collection  are  four  specimens 
from  the  Hauterivian  of  Oued  Tidzi,  one  from  the  Hauterivian  of  Ifrech  Oued  Igoulouzen,  four  from 
the  Barremian  of  Ait  el  Faci,  and  seven  from  the  Barremian  of  Asif  Ait  Ameur  which  are  probably 
this  species.  M.  moutoniana  is  a sub-Tethyan  species  of  very  widespread  occurrence  throughout  the 
Lower  Cretaceous  (see  Middlemiss  1976,  1979)  and  it  would  indeed  be  surprising  if  some  specimens 
were  not  to  be  found  in  south-west  Morocco. 


EXPLANATION  OF  PLATE  61 

Fig.  1 . Loriolithyris  melaitensis  sp.  nov.  Section  6.8  of  specimen  S. 556/1  (not  included  in  text-fig.  12)  enlarged 
to  show  the  development  of  the  crural  base  with  secondary  clubbing.  The  primary  hinge  plate  has  a cuneate 
relationship  to  the  crural  base. 

Fig.  2.  Loriolithyris  melaitensis  sp.  nov.  Section  4.8  of  text-fig.  1 2 photographed  to  show  the  internal  structure 
of  the  cardinal  process,  especially  the  distribution  of  punctate  and  impunctate  skeletal  tissue.  The  juvenile 
primary  hinge  plates  have  a secondary  clubbed  thickening  which  was  deposited  prior  to  the  incorporation  of 
the  hinge  plates  into  the  cardinal  process. 

Fig.  3.  Loriolithyris  melaitensis  sp.  nov.  Section  5.2  of  text-fig.  12  enlarged  to  show  the  primary  hinge  plate 
surrounded  by  secondary  tissue  and  the  first  sign  of  development  of  the  crural  base  within  the  piped  inner 
margin  of  the  hinge  plate. 

Fig.  4.  Loriolithyris  russillensis  (de  Loriol).  Section  4.6  of  text-fig.  7 enlarged  to  show  the  structure  of  the  piped 
inner  margin  of  the  hinge  plate. 

Fig.  5.  Paraboubeithyris  plicae  gen.  et  sp.  nov.  Part  of  section  4.8  of  text-fig.  16  enlarged  to  show  the  structure 
of  the  corniced  inner  margin  of  the  hinge  plate. 

Linear  scale  = 2 mm. 


PLATE  61 


middlemiss,  Cretaceous  Terebratulidae 


554 


PALAEONTOLOGY,  VOLUME  23 


Sellithyris  carteroniana  (d’Orbigny)  is  recorded  by  Roch  from  the  Berriasian  and  the  Barremian, 
by  Gigout  from  the  Valanginian  (of  Safi)  and  by  Ambroggi  from  the  Hauterivian.  In  the  Gentil 
Collection  there  is  one  specimen  from  Tinirt  Ait  Ameur  (probably  Hauterivian)  which  has  some 
resemblance  to  S.  carteroniana  in  being  obese,  equidimensional,  and  strongly  episulcate  but  the 
resemblance  is  closer,  in  fact,  to  the  Algerian  variety  or  subspecies  of  S.  sella  (see  below).  The  same 
can  be  said  of  Gigout’s  figured  specimen  (Gigout  1951,  pi.  9,  figs.  11-14).  S.  carteroniana  is  an 
interesting  species  from  the  palaeobiogeographical  point  of  view  as  (a)  it  is  a characteristic  member  of 
the  Jura  fauna  which  is  also  found  in  north  Germany  during  the  time  of  the  Valanginian-Hauterivian 
transgression  (Middlemiss  1976,  1979)  and  ( b ) Terebratula  coahuilensis  of  the  Neocomian  of 
northern  Mexico  is  probably  synonymous  with  it.  In  view  of  my  thesis  of  the  Jura  affinities  of  the 
south-west  Moroccan  fauna  the  occurrence  of  this  species  would  be  significant.  Unfortunately  there 
is  no  evidence  that  all  the  records  do  not  refer  to  S.  sella,  although  some  may  refer  to  Boubeithyris 
pleta. 

Sellithyris  sella  (J.  de  C.  Sow)  is  recorded  by  both  Roch  and  Ambroggi  from  the  Barremian  and 
Gargasian  and  by  Roch  from  the  Bedoulian  also.  This  almost  ubiquitous  Lower  Cretaceous  species 
would  be  expected  to  occur  in  south-west  Morocco,  especially  as  an  undescribed  form  of  it  is 
certainly  known  from  the  Lower  Cretaceous  of  the  High  Plateaux  region  of  Algeria.  In  the  Gentil 
Collection  are  twenty-three  specimens  from  Tinirt  Ait  Ameur  (labelled  Berrisian-Valanginian  but 
more  likely  Hauterivian)  which  appear  to  be  this  obese  Algerian  variety  of  the  species.  There  is  also 
one  specimen  from  the  Hauterivian  of  Oued  Tidzi,  one  from  the  Barremian  of  Ida  ou  Troumma,  and 
two  from  the  Barremian  of  Tibourr’m;  these  resemble  the  more  normal  somewhat  depressed 
Neocomian  form  of  the  species. 

Moutonithyris  dutempleana  (d’Orbigny).  This  almost  ubiquitous  Albian  species  is  recorded  by  both 
Roch  and  Ambroggi  from  both  the  Clansayesian  and  the  Albian.  Its  occurrence  in  the  Albian  would 
not  be  surprising.  Doubts  are  raised,  however,  by  two  circumstances:  (a)  M.  dutempleana  is  very  rare 
in  the  Clansayesian  and  known  certainly  from  that  stage  only  in  Sardinia  (Dieni  et  al.  1975).  On  the 
other  hand  if,  as  is  likely,  the  species  spread  from  south  to  north,  it  could  well  occur  in  the 
Clansayesian  of  Morocco.  ( b ) Cyrtothyris  middlemissi  certainly  occurs  in  both  Clansayesian  and 
Albian  and  is  easily  mistaken  for  M.  dutempleana  (Calzada  1972,  p.  66).  The  specimen  figured  by 
Gigout  (1951,  pi.  13,  figs.  5-8)  as  T.  biplicata  is  a Concinnithyris  cf.  obesa. 

To  summarize:  T.  sueuri,  T.  collinaria,  T.  salevensis,  T.  carteroniana,  and  M.  dutempleana  have 
probably  been  misidentified  by  previous  authors.  M.  moutoniana  and  S.  sella  probably  do  occur 
rarely  in  south-west  Morocco. 


Acknowledgements.  I particularly  thank  Mr.  E.  F.  Owen  (British  Museum,  Natural  History),  Monsieur 
D.  Pajaud  and  his  staff  at  the  Universite  Pierre  et  Marie  Curie,  and  Monsieur  J-P.  Thieuloy  (Grenoble).  I thank 
Mademoiselle  S.  Willefert  (Service  de  la  Carte  Geologique  du  Maroc,  Rabat)  for  her  help  in  locating  and  sending 
to  me  specimens  from  the  Roch  Collection;  also  the  Head  of  the  Laboratoire  de  Geologie-Paleontologie, 
Universite  Mohamed  V,  Rabat,  and  Monsieur  G.  Cogne  for  lending  me  specimens  figured  by  Gigout. 
Additional  specimens  were  lent  by:  Professor  D.  V.  Ager  (Swansea),  Senor  S.  Calzada  Badia  (Barcelona), 
Monsieur  Y.  Champetier  (Nancy),  Monsieur  A.  Charriere  (Paris),  Monsieur  M.  Debuyser  (Paris),  Professor 
M.  Durand  Delga  (Toulouse),  Dr.  W.  J.  Kennedy  (Oxford),  Monsieur  E.  Lanterno  (Geneva),  Monsieur 
Y.  Rangheard  (Besan9on),  Dr.  J.  Remane  (Neuchatel),  and  Monsieur  Weidmann  (Lausanne). 


REFERENCES 

ager,  d.  v.  1974.  The  western  High  Atlas  of  Morocco  and  their  significance  in  the  history  of  the  North  Atlantic. 
Proc.  geol.  Ass.,  Lond.  85,  23-41,  London. 

— and  evamy,  b.  D.  1964.  The  geology  of  the  southern  French  Jura.  Ibid.  74  (for  1963),  325-355,  pi.  9, 
London. 

ambroggi,  r.  1963.  Etude  geologique  du  versant  meridionel  du  Haut  Atlas  occidental  et  de  la  Plaine  du  Souss. 
Notes  et  Mem.  Serv.  geol.  Maroc,  157,  321  pp.,  181  figs.,  Rabat. 


MIDDLEMISS:  CRETACEOUS  TEREBRATULIDAE 


555 


arkell,  w.  J.  1956.  Jurassic  Geology  of  the  World , Edinburgh. 

barczyk,  w.  1969.  Upper  Jurassic  terebratulids  from  the  Mesozoic  border  of  the  Holy  Cross  Mountains  in 
Poland.  Pr.  Muz.  Ziemi,  14,  1-82,  pis.  1-18,  Warsaw. 

bogdanova,  t.  n.  and  lobacheva,  s.  v.  1966.  Neocomian  Fauna  of  the  Kopet-Daga.  Min.  Geol.  U.S.S.R.,  Inst. 

Econ.  Sci.  Leningrad,  n.s.  130,  1-140,  pis.  1-13,  Leningrad.  [In  Russian.] 
bonneau,  m.,  beauvais,  l.  and  middlemiss,  f.  a.  1975.  L’unite  de  Miamou  (Crete-Grece)  et  sa  macrofaune  d’age 
Jurassique  superieur  (Brachiopodes,  Madreporaires).  Ann.  Soc.  geol.  Nord,  94  (for  1974),  71-85,  pis.  10-11, 
Lille. 

boullier,  A.  1976.  Les  terebratulides  de  VOxfordien  du  Jura  et  de  la  bordure  sud  du  Bassin  de  Paris.  Thesis, 
Besangon. 

buckman,  s.  s.  1918.  The  Brachiopods  of  the  Namyau  Beds,  northern  Shan  States,  Burma.  Pal.  Indica,  n.s.  3 
(for  1917),  Mem.  no.  2,  1-254,  pis.  1-21,  Calcutta. 

busnardo,  R.  and  durand-delga,  m.  1960.  Donnees  nouvelles  sur  le  Jurassique  et  le  Cretace  inferieur 
dans  Test  des  Cordilleres  Betiques  (Regions  d’Alcoy  et  d’ Alicante).  Bull.  Soc.  geol.  Fr.  (7),  2,  278-287, 
Paris. 

calzada,  s.  1972.  Cyrtothyris  middlemissi,  n.  sp.  del  Aptiense  de  Garraf  (Barcelona).  Acta  geol.  Hisp.  7,  66-68, 
2 figs.,  Barcelona. 

choubert,  g.,  faure-muret,  a.  and  hottinger,  l.  1967.  Apergu  geologique  du  bassin  cotier  de  Tarfaya.  Notes 
Mem.  Serv.  geol.  Maroc,  175  (1),  7-106,  Rabat. 

cox,  Margaret  m.  and  middlemiss,  f.  a.  1978.  Terebratulacea  from  the  Cretaceous  Shenley  Limestone. 

Palaeontology,  21,  411-441,  pis.  40-42,  figs.  1-13,  London. 
d’archiac,  a.  1847.  Rapport  sur  les  fossiles  du  Tourtia.  Mem.  Soc.  geol.  Fr.,  (2)  2,  291-351,  pis.  13-25,  Paris. 
dieni,  I.,  middlemiss,  f.  a.  and  owen,  e.  f.  1975.  The  Lower  Cretaceous  Brachiopods  of  east-central  Sardinia. 

Bol.  Soc.  Pal.  Ital.  12  (for  1973),  166-216,  pis.  32-38,  Modena. 
geyssant,  jeannine.  1966.  Glossothyris  et  Pygope  (Terebratulidae)— essai  de  repartition  de  ces  especes  dans  la 
domaine  mediterraneen.  Notes  Serv.  geol.  Maroc,  26,  75-98,  pis.  1-3,  figs.  1-7,  8 tables,  Rabat. 
gignoux,  m.  1955.  Stratigraphic  Geology  (English  edn.),  San  Francisco. 

gigout,  m.  1951.  Etudes  geologiques  sur  la  Meseta  marocaine  occidentale  (arriere-pays  de  Casablanca, 
Mazagan  et  Safi).  Notes  Mem.  Serv.  geol.  Maroc,  86,  1-507,  pis.  1-18,  Rabat. 
imlay,  R.  w.  1937.  Lower  Neocomian  fossils  from  the  Miquihuana  region,  Mexico.  J.  Palaeont.  11,  552-574, 
pis.  70-83,  Menasha. 

— 1940.  Neocomian  faunas  of  northern  Mexico.  Bull.  G.  S.  Amer.  51,  117-190,  pis.  1-21,  New  York. 
kyansep,  n.  p.  1961 . Terebratulids  of  the  Lusitanian  Beds  of  the  Lower  Kimmeridgian  of  the  south-west  Crimea. 

Akad.  Nauk.  U.S.S.R.  8,  1-101,  8 pis.,  Moscow.  [In  Russian.] 
loriol,  p.  de.  1866.  Description  de  fossiles  de  V oolite  corallienne,  de  Vetage  valangien  et  de  Vetage  urgonien  du 
Mont  Saleve,  Geneva. 

1867.  In  favre,  A.  Recherches  geologiques  dans  la  Savoie,  Paris  and  Geneva. 

— 1868.  Monographie  des  couches  de  l’etage  valangien  des  Carrieres  d’Arzier  (Vaud).  Mater,  pour  Paleont. 
suisse,  ser.  4,  Geneva. 

— and  gillieron,  v.  1869.  Monographie  paleontologique  et  stratigraphique  de  l’etage  urgonien  inferieur  du 
Landeron  (Neuchatel).  Mem.  Soc.  helv.  Sc.  nat.  23  (5),  1-123,  pis.  1-18,  Zurich. 

middlemiss,  F.  A.  1959.  English  Aptian  Terebratulidae.  Palaeontology,  2,  94-142,  pis.  15-18,  London. 

1968a.  Brachiopodes  du  Cretace  inferieur  des  Corbieres  orientales  (Aude).  Ann.  Paleont.  (Invert.),  54, 

173-197,  pis.  A-C,  Paris. 

1968 b.  Observations  on  the  ontogeny  of  the  brachiopod  Sellithyris  sella.  Bull.  Ind.  geol.  Ass.  1,  1-17,  pi.  1, 

Chandigarh. 

1973.  The  geographical  distribution  of  Lower  Cretaceous  Terebratulacea  in  western  Europe.  In  casey,  r. 

and  rawson,  p.  f.  (eds.).  The  boreal  Lower  Cretaceous:  Geol.  J.  Spec.  Issue  no.  5,  1 10-129,  Liverpool. 

— 1976.  Lower  Cretaceous  Terebratulidina  of  northern  England  and  Germany  and  their  geological  back- 
ground. Geol.  Jb.  A30,  21-104,  pis.  1-11,  Hanover. 

— 1979.  Boreal  and  Tethyan  brachiopods  in  the  European  early  and  middle  Cretaceous.  Kreide  Europas 
IUGS  ser.  A. 

muir-wood,  h.  M.  1965.  In  moore,  R.  c.  (ed.).  Treatise  on  Invertebrate  Paleontology,  pt.  H,  New  York. 

Pictet,  f-j.  and  loriol,  p.  de.  1 872.  Description  des  fossiles  du  terrain  cretace  des  environs  de  Sainte-Croix, 
pt.  5.  Mater. pour  la  Paleont.  suisse  (6),  Geneva. 

ROCH,  E.  1930.  Etudes  geologiques  dans  la  region  meridionale  du  Maroc  occidentale.  Notes  et  Mem.  Serv.  Mines 
Maroc,  9,  1-542,  pis.  1-26,  Macon. 


556 


PALAEONTOLOGY,  VOLUME  23 


Smirnova,  t.  n.  1960.  Brachiopoda.  In  drushchitz,  v.  v.  and  kudriavcheva,  m.  p.  Atlas  of  the  Lower  Cretaceous 
Fauna  of  the  northern  Caucasus  and  Crimea.  Trudy,  VNII  Gaz.,  370-387,  6 pis.,  Moscow.  [In  Russian.] 

— 1969.  A new  terebratulid  genus  from  the  Tithonian-Valanginian.  Pal.  Zh.,  3,  144-146,  Moscow.  [In 
Russian.] 

— 1972.  Brachiopods  from  the  Crimea  and  northern  Caucasus.  Akad.  Nauk.  U.S.S.R.  140  pp.,  13  pis., 
Moscow.  [In  Russian.] 


Typescript  received  5 March  1979 

Revised  typescript  received  15  November  1979 


F.  A.  MIDDLEMISS 

Department  of  Geology 
Queen  Mary  College 
Mile  End  Road 
London 


COLLIGNONICERATID  AMMONITES  FROM 
THE  MID-TURONIAN  OF  ENGLAND  AND 
NORTHERN  FRANCE 

by  W.  J.  KENNEDY,  C.  W.  WRIGHT,  and  J.  M.  HANCOCK 


Abstract.  Collignoniceras  Breistroffer,  1947  is  represented  by  five  species  in  the  mid-Turonian  of  England  and 
Touraine  (the  type  area  of  the  Turonian  stage)  in  northern  France.  The  cosmopolitan  and  highly  variable  type 
species  C.  woollgari  (Mantell)  is  shown  to  be  a senior  synonym  of  C.  schlueterianum  (Laube  and  Bruder)  and 
C.  mexicanum  (Bose)  amongst  others,  and  shows  features  indicating  that  Selwynoceras  Warren  and  Stelck,  1940 
(the  type  species  of  which  S.  boreale  (Warren),  is  also  redescribed)  is  a synonym  of  Collignoniceras  sensu  stricto. 
Other  species  referred  to  the  genus  are  C.  carolinum  (d’Orbigny),  C.  papale  (d’Orbigny),  C.  canthus  (Sornay)  and 
C.  turoniense  (Sornay).  Ammonites  fleuriausianus  d’Orbigny,  1841  is  a senior  synonym  of  A.  vielbancii  d’Orbigny, 
1850  and  is  made  the  type  species  of  Lecointriceras  gen.  nov.,  to  which  two  further  species,  L.  carinatum  sp.  nov. 
and  L.  costatum  sp.  nov.  are  also  referred. 

Collignoniceras  woollgari  (Mantell)  is  one  of  the  most  widely  cited  mid-Cretaceous  ammonite  species, 
giving  its  name  to  the  middle  zone  of  the  Turonian  standard  sequence  (Wright  in  Arkell  et  al.  1957; 
Rawson  et  al.  1978).  As  with  other  classic  species,  the  type  material  has  never  been  adequately  figured 
and  is  of  uncertain  horizon,  although  it  has  at  least  survived  the  vicissitudes  of  a century  and  a half 
since  its  original  description  (Mantell  1822,  p.  197;  pi.  21,  fig.  16;  pi.  22,  fig.  7).  In  England,  where  it 
was  first  described,  the  species  is  rare  and  the  lectotype  remains  the  only  good  adult  specimen  known. 
Elsewhere,  however,  it  is  recorded  abundantly,  especially  in  the  U.S.  Western  Interior  region,  where 
it  formed  the  basis  of  one  of  the  early  accounts  of  intraspecific  variability  in  Cretaceous  ammonites 
(Haas  1946),  although  as  Haas  and  Meek  before  him  (1876,  p.  455)  noted,  authors  have  questioned 
whether  the  great  majority  of  specimens  referred  to  this  cosmopolitan  species  are  indeed  conspecific 
with  Mantell’s  types. 

We  have  studied  hundreds  of  European,  American  and  Japanese  Collignoniceras  in  connection 
with  this  project,  and  encountered  an  initially  bewildering  range  of  variation,  both  in  adult  ornament 
and  the  size  at  which  ontogenetic  changes  occur.  We  have  relatively  few  juveniles  from  Europe  but 
many  from  the  U.S. A.;  conversely,  large  complete  adults  are  common  in  European  collections,  but 
those  from  the  U.S.  are  usually  fragmentary.  Whilst  it  would  be  possible  to  select  individuals  with 
differences  that  could  be  framed  into  diagnostic  features  for  specific  or  subspecific  separation,  this 
would  be  misleading  and  conceal  the  over-all  common  features  of  the  species  recognized  below.  In 
C.  woollgari  in  particular  we  have  no  doubt  that  a series  of  local  races  of  the  species  existed  over  its 
wide  spread,  but  to  separate  formally  the  successive  or  local  populations,  differing  in  the  extent  of 
morphological  variation  but  overlapping,  would  serve  no  useful  purpose.  The  broad,  variable  species 
described  below  not  only  represent  reality  but  are  adequate  for  detailed  correlation  and  discussion  of 
the  evolution  of  the  genus. 


SYSTEMATIC  DESCRIPTIONS 

Location  of  specimens.  The  following  abbreviations  are  used  to  indicate  the  repositories  of  specimens  studied: 
AM  Museum  de  Paleontologie  d’Angers. 

BMNH  British  Museum  (Natural  History),  London. 

CS  Chateau  de  Saumur 

EMP  Ecole  des  Mines,  Paris  (now  housed  at  the  Universite  Claude  Bernard,  Lyon). 

IPalaeontology,  Vol.  23,  Part  3,  1980,  pp.  557-603,  pis.  62-77.| 


558 


PALAEONTOLOGY,  VOLUME  23 


FSM  Faculte  des  Sciences,  Le  Mans;  chiefly  collections  formerly  housed  in  the  Musee  de  Tesse,  Le  Mans. 
FSR  Institut  de  Geologie,  Universite  de  Rennes. 

GK  Department  of  Geology,  Kyushu  University,  Fukuoka. 

MNHP  Museum  National  d’Histoire  Naturelle,  Paris. 

OUM  University  Museum,  Oxford;  unless  stated  otherwise,  these  are  collections  made  by  Hancock  and 
Kennedy. 

SP  Collections  of  the  Sorbonne,  now  Universite  de  Paris  VI. 

WW  C.  W.  and  E.  V.  Wright  collection. 

Dimensions.  All  dimensions  are  given  in  millimetres;  figures  in  parentheses  are  the  dimensions  as  a percentage  of 
the  total  diameter.  D = diameter;  Wb  = whorl  breadth;  Wh  = whorl  height;  U = umbilicus;  Ic  = intercostal; 
c = costal;  R = number  of  ribs  per  whorl. 

Suture  terminology.  The  suture  terminology  of  Wedekind  (1916;  see  Kullman  and  Wiedmann  1970  for  a recent 
review)  is  followed  here:  I = Internal  lobe,  U = Umbilical  lobe,  L = Lateral  lobe,  E = External  lobe. 


Suborder  ammonitina  Hyatt,  1889 
Superfamily  acanthocerataceae  de  Grossouvre,  1 894 
[nom  transl.  et  correct.  Hyatt  1900,  ex  Acanthoceratides  de  Grossouvre,  1894] 

Family  collignoniceratidae  Wright  and  Wright,  1951 
Subfamily  collignoniceratinae  Wright  and  Wright,  1951 
Genus  collignoniceras  Breistroffer,  1947 
{non  Van  Hoepen,  1955) 

Type  species.  Ammonites  woollgari  Mantell,  1822  by  the  original  designation  of  Meek  (1876)  as  type  species  of 
Prionotropis  Meek,  1876  {non  Fieber,  1853),  for  which  Breistroffer  (1947)  proposed  Collignoniceras  as  nomen 
novum. 

Diagnosis.  Medium  to  large,  moderately  involute  to  evolute  ammonites.  Early  whorls  compressed, 
parallel  sided,  ornamented  by  crowded  or  sparse,  prorsiradiate,  straight  or  flexuous  ribs,  mostly  long, 
with  weak  to  strong  umbilical  bullae.  All  ribs  bear  in  the  early  stages  outer  ventrolateral  tubercles  in 
addition  to  siphonal  clavi. 

This  style  of  ornament  is,  in  some  species,  retained  to  maturity.  In  most,  however,  the  ribs  coarsen, 
become  widely  spaced,  with  strong  to  weak  umbilical  tubercles  (which  migrate  progressively 
outwards  from  the  umbilical  margin),  prominent  inner  and  outer  ventrolateral  tubercles  which  may 
fuse  into  a massive  horn  or  flared  rib,  from  which  commonly  arise  pairs  of  low  ribs,  joining  siphonal 
clavi  more  numerous  than  the  ventrolateral  and  linked  into  a more  or  less  continuous  keel.  Rarely  the 
ornament  is  greatly  reduced  on  the  body  whorl. 

The  sutures  are  little  incised,  with  massive  saddles. 

Discussion.  The  diagnosis  given  above  summarizes  the  rather  wide  variation  seen  in  species  referred 
to  this  genus,  which  include  C.  boreale  (Warren),  C.  papale  (d’Orbigny),  C.  canthus  (Sornay), 
C.  turoniense  (Sornay)  and  C.  carolinum  (d’Orbigny).  The  nomenclatorial  history  of  the  genus  is 
somewhat  complex.  Meek  introduced  a subgenus  Prionotropis  in  1876,  with  Ammonites  woollgari 
Mantell  as  type  species.  Breistroffer  (1947)  pointed  out  the  prior  usage  of  Prionotropis  by  Fieber 
(1853)  and  proposed  Collignoniceras  as  nomen  novum.  Meanwhile  Warren  and  Stelck  (1940)  had 
proposed  the  genus  Selwynoceras  with  P.  borealis  Warren,  1930  as  type  species,  distinguishing  it  from 
Meek’s  Prionotropis  by  the  presence  of  a row  of  nodes  instead  of  a keel  on  the  inner  whorls  and  the 
marked  alternation  in  length  and  strength  of  the  ribs.  Wright  (in  Arkell  et  al.  1957,  p.  L426)  regarded 
Selwynoceras  as  a subgenus  of  Collignoniceras , whilst  Powell  (1963,  p.  1223)  considered  the  two 
synonymous.  Following  an  application  by  Matsumoto  and  Wright  in  1966,  the  International 
Commission  on  Zoological  Nomenclature  ruled  in  1968  (Opinion  861)  that  Collignoniceras 
Breistroffer,  1947,  should  be  given  priority  over  Selwynoceras  Warren  and  Stelck,  1940,  by  those 
who  regard  the  two  as  synonyms. 


KENNEDY  ET  AL..  COLLIGNONICER ATI D AMMONITES 


559 


From  a comparison  of  the  types  and  other  specimens  of  C.  woollgari  and  S.  bore  ale,  we  would  agree 
with  Powell  that  the  two  do  not  bear  even  subgeneric  separation:  boreale  is  simply  a small  species  of 
Collignoniceras  in  which  the  flared  ribs  appear  at  a relatively  early  stage.  The  ventral  tuberculation 
visible  on  the  outer  whorl  of  the  lectotype  (here  designated),  which  is  refigured  here  as  PI.  70,  figs.  1 -2, 
is  on  exactly  the  same  plan  as  in  English  woollgari,  whilst,  as  Haas  (1946),  Powell  (1963)  and 
Matsumoto  (1965)  have  shown,  the  style  of  ribbing  of  juvenile  Collignoniceras  is  very  variable. 

Collignoniceras  differs  from  Prionocyclus  Meek,  1876  (type  species  P.  wyomingensis  Meek)  in  that 
the  latter  has  very  fine  dense  irregular  ribs  through  most  or  all  of  its  ontogeny  and  a broader  venter 
with  an  entire  or  serrated  keel.  C.  woollgari  and  P.  hyatti  (Stanton)  overlap  in  time  in  the  southern 
U.S.  and  some  late  C.  woollgari  there  and  also  in  Europe  show  a low  siphonal  keel  at  maturity, 
emphasizing  the  intimate  relationship  between  the  two.  Ribbing  is  usually  dominant  over 
tuberculation  in  Prionocyclus,  although  some  species  bear  finger-like  ventrolateral  spines,  fore- 
shadowing the  development  seen  in  the  later  Prionocycloceras  (Young  1963,  pi.  23,  figs.  1-6;  pi.  27, 
figs.  2-4).  Matsumoto  (1965,  p.  19)  discusses  other  differences  between  these  two  genera. 

Subprionocyclus  Shimizu,  1932  was  originally  separated  from  Collignoniceras  [Prionotropis]  on  the 
basis  of  minor  differences  between  the  internal  sutures.  As  Matsumoto  (1959,  p.  109)  notes,  however, 
distinguishing  features  also  include  the  paired  or  alternately  long  and  short  sigmoidal  ribs  of 
Subprionocyclus  which  may  flatten  on  the  outer  whorl,  greater  persistence  of  outer  ventrolateral 
tubercles  and  absence  of  massive  horns.  Like  Prionocyclus,  Subprionocyclus  has  a continuous 
persistent  keel  which  varies  with  the  density  of  the  ribbing  from  finely  to  coarsely  serrate. 

Germariceras  Breistroffer,  1947  is  perhaps  only  doubtfully  separable  from  Prionocyclus',  known 
only  from  juveniles,  it  may  be  separated  from  Collignoniceras  by  the  possession  of  fine  dense  narrow 
ribs  with  small  sharp  umbilical,  inner  and  outer  ventrolateral  tubercles  and  a finely  serrated 
continuous  keel  with  more  serrations  than  the  number  of  ventrolateral  tubercles. 

Reesidites  Wright  and  Matsumoto,  1964,  which  should  perhaps  be  placed  in  Barroisiceratinae,  is 
compressed  and  involute,  high  whorled,  with  a fastigiate  venter;  sinuous  ribs  branch  in  groups  of  two 
or  three  from  small  umbilical  bullae,  with  single  ventrolateral  and  siphonal  clavi  only.  The  largest 
individuals  barely  exceed  100  mm  diameter  (Matsumoto  1965). 

Subprionotropis  Basse,  1950,  known  only  from  specimens  a few  centimetres  in  diameter,  differs 
from  Collignoniceras  in  being  involute  with  compressed  whorls,  with  ribs  arising  in  pairs  from 
umbilical  bullae  (with  additional  intercalated  ribs)  bearing  only  ventrolateral  and  siphonal  clavi  and 
forming  strong  chevrons  on  the  fastigiate  venter.  At  the  end  of  the  body  chamber,  ribs  and  tubercles 
weaken  and  the  venter  becomes  rounded. 

Lymaniceras  Matsumoto,  1965  and  Niceforoceras  Basse,  1948  are  both  compressed  and  involute, 
with  weak,  dense  flexuous  ribs  or  striae,  a single  ventrolateral  tubercle  and  a finely  serrated  keel. 

Collignoniceras  is  the  earliest  genus  of  Collignoniceratidae  to  appear  in  the  Turonian,  and,  as 
Matsumoto  (1965)  has  noted,  some  individuals  in  variable  United  States  Western  Interior 
populations  show  early  whorls  which  foreshadow  Prionocyclus,  Subprionocyclus  and  thence  the 
remaining  late  members  of  the  group. 

With  respect  to  the  evolutionary  origins  of  the  genus,  Wright  (in  Arkell  et  al.  1957,  p.  L426)  and 
Matsumoto  (1965,  p.  10)  have  suggested  that  the  diminutive  late  Cenomanian  acanthoceratinid 
Protacanthoceras  Spath,  1923  might  be  the  ultimate  ancestor,  with  Neocar diocer as  Spath,  1926  as  an 
intermediate.  Recent  collecting  from  the  latest  Cenomanian/early  Turonian  faunas  of  the  condensed 
Neocar  diocer  as  Pebble  Bed  of  Devon  (see  Hancock,  Kennedy  and  Wright  1977,  fig.  2 for  details)  has 
now  produced  a range  of  specimens  that  provisionally  we  refer  to  Thomelites  Wright  and  Kennedy, 
1973,  among  which  are  individuals  with  siphonal  clavi  tending  to  form  a continuous  serrated  keel.  In 
addition,  a few  poorly  preserved  fragments  seem  already  to  have  reached  the  stage  of  Collignoniceras 
in  some  features  of  decoration.  There  remains,  however,  a gap  in  the  European  successions, 
corresponding  to  most  of  the  Mammites  nodosoides  assemblage  Zone,  in  which  the  genus  is  absent 
apart  from  a single  possible  example  in  the  collection  of  Colonel  O.  H.  Bayliss,  from  Shapwick, 
Devon;  W.  A.  Cobban  (in  litt.,  1978)  tells  us  that  Collignoniceras  appears  at  the  top  of  the  North 
American  correlatives  of  this  zone. 


560 


PALAEONTOLOGY,  VOLUME  23 


Occurrence.  Collignoniceras  is  widespread  in  the  middle  of  the  Turonian  stage,  the  classic  woollgari  Zone.  There 
are  records  from  England,  France,  Germany,  Czechoslovakia,  Poland,  Rumania,  Turkestan,  Japan,  California, 
Texas,  the  U.S.  and  Canadian  Interiors,  Greenland,  north  Africa,  Colombia,  and  northern  Australia. 


Collignoniceras  woollgari  (Mantell) 


Plates  62-67 ; Plate  69,  figs.  3-4;  Plate  71,  figs.  1-3;  text-figs.  1a,  2-4 


1822 
non  1841 
1850 
1855 
1860 
1867 
1872 
1872 
1872 


1887 

1887 

1902 

1902 

1902 

1907 

1925 

1928 

1931 

1931 

1931 

1946 

1963 

1971 

1972 

1975 

1977 

1977 


Ammonites  Woollgari  Mantell,  p.  197,  pi.  21,  fig.  16;  pi.  22,  fig.  7. 

Ammonites  Woollgari  Mantell;  d’Orbigny,  p.  352,  pi.  108,  figs.  1-3. 

Ammonites  Woolgarii  d’Orbigny,  p.  189  (pars). 

Ammonites  Woollgari  Mantell;  Sharpe,  p.  27,  pi.  11,  figs.  1,  2. 

Ammonites  carolinus  (d’Orbigny);  Courtiller,  p.  251,  pi.  3,  fig.  2. 

Ammonites  Woolgarii  Mantell;  Courtiller,  p.  7,  pi.  8,  figs.  1-4. 

Ammonites  Woolgari  Mantell;  Schliiter,  p.  25,  pi.  9,  figs.  1-5;  non  pi.  12,  figs.  5,  6. 
Ammonites  Woollgari  Mantell;  Geinitz,  p.  184,  pi.  33,  figs.  1,  2 (?),  non  4-5. 

Ammonites  Woolgari  Fritsch,  p.  30  (pars),  pi.  3,  figs.  1-3;  pi.  4,  figs.  1-2;  pi.  14,  fig.  6;  non  pi.  2, 
tigs.  1-2;  pi.  15,  fig.  6. 

Acanthoceras  Woollgari  (Mantell);  Laube  and  Bruder,  p.  235,  text-fig. 

Acanthoceras  Schliiterianum  Laube  and  Bruder,  p.  236,  pi.  29,  figs.  2-3. 

Acanthoceras  Woollgari  (Mantell);  Petrascheck,  p.  149,  text-figs.  7-8. 

Acanthoceras  cfr.  Woollgari  (Mantell);  Petrascheck,  p.  148,  pi.  12,  figs.  2-3. 

Acanthoceras  Schliiterianum  Laube  and  Bruder;  Petrascheck,  p.  150,  pi.  10,  fig.  3;  pi.  11,  fig.  3; 
pi.  12,  fig.  1. 

Prionotropis  Schliiterianum  Laube  and  Bruder;  Pervinquiere,  p.  275. 

Prionotropis  Schliiteriana  Laube  and  Bruder;  Diener,  p.  156. 

Prionotropis  woollgari  Mantell  var.  mexicana  Bose,  p.  262,  pi.  1 1,  figs.  11,  12. 
Pseudaspidoceras(l)  chispaense  Adkins,  p.  51,  pi.  3,  figs.  1-2. 

Pseudaspidocerasl  sp.  Adkins,  p.  53,  pi.  2,  fig.  2. 

Pseudaspidoceras(l)  n.sp.  A;  Adkins,  p.  53,  pi.  3,  figs.  3-4. 

Prionotropis  woollgari  Meek  (?  non  Mantell);  Haas,  p.  150,  pis.  11,  12;  pi.  13,  figs.  1-3,  5-18; 
pi.  14,  figs.  1-10,  12-16;  pi.  15,  figs.  1-6,  9,  10;  pis.  16,  17;  pi.  18,  figs.  1-2,  7-9;  text-figs.  1-91. 
Selwynoceras  mexicanum  (Bose);  Powell,  p.  1225,  pi.  166,  figs.  2-7;  pi.  167,  figs.  1,  3-8;  pi.  168, 
fig.  4;  text-figs.  2-4. 

Collignoniceras  woollgari  (Mantell);  Matsumoto,  p.  130,  pi.  21,  fig.  4,  text-fig.  1. 
Collignoniceras  woollgari  (Mantell);  Cobban  and  Scott,  p.  94,  pi.  14,  fig.  5;  pi.  30,  fig.  1;  pi.  37, 
figs.  9-10  (with  additional  synonymy). 

Collignoniceras  woollgari  (Mantell);  Hattin,  pi.  10,  figs.  N,  P,  Q,  R. 

Collignoniceras  (Selwynoceras)  schlueterianum  (Laube  and  Bruder);  Hancock,  Kennedy  and 
Wright,  p.  156. 

Collignoniceras  (Collignoniceras)  cf.  C.  woollgari  sensu  Matsumoto,  1965,  group  E;  Hancock, 
Kennedy  and  Wright,  p.  156. 


Types.  The  lectotype,  designated  by  Wright  and  Wright  (1951,  p.  35),  is  BMNH  5682,  from  the  Middle  Chalk  of 
Lewes,  Sussex,  refigured  here  as  Plate  62,  figs.  1-2;  Plate  63,  fig.  9.  Two  additional  specimens  from  Mantell’s 
collection,  BMNH  C5742  a-b  (Plate  69,  figs.  3,  4),  are  presumed  to  be  paralectotypes. 


Other  specimens  studied.  These  include:  BMNH  4863  a-b,  from  the  Middle  Chalk  ‘near  Lewes,  Sussex’;  BMNH 
43963  ‘Lower  Chalk,  near  Lewes’  (J.  de  C.  Sowerby  Collection);  BMNH  C30394  ‘Turonian  Mount  Caburn  Pit, 
near  Glynde,  Sussex’  (labelled  aff.  woollgari  by  L.  F.  Spath);  BMNH  C40152  from  the  Middle  Chalk, 
Terebratulina  lata  Zone,  Mickleham  Bypass,  Surrey  (C.  W.  and  E.  V.  Wright  Collection);  WW  16682,  14792-4, 
from  the  Middle  Chalk,  top  of  the  T.  lata  Zone  Middle  Chalk,  Mickleham  Bypass,  Surrey;  WW  22925-7,  Middle 


EXPLANATION  OF  PLATE  62 

Figs.  1 -2.  Collignoniceras  woollgari  (Mantell).  The  lectotype,  BMNH  5682,  from  the  Middle  Chalk  of  Lewes, 
Sussex. 


PLATE  62 


KENNEDY,  wright  and  Hancock,  Collignoniceratid  ammonites 


562 


PALAEONTOLOGY,  VOLUME  23 


Chalk,  Lewknor  Crossroads,  Lewknor,  Oxon.  (ex  R.  E.  H.  Reid  Collection);  OUM  K 10273,  K 10275-76  from 
no  more  than  5 m below  the  top  of  the  Chalk  Rock  at  Fognam  Barn,  Berkshire,  3 km  WNW  of  Lambourn. 

BMNH  88988  b,  88989  a-c  from  the  Turonian  of  the  White  Mountain,  near  Prague,  Czechoslovakia. 

French  specimens  include  the  following:  OUM  KZ  741,  743-4,  746,  748-9,  753,  from  the  St.  Cyr-en-Bourg 
Fossil  Bed,  Champignonniere  Les  Rochains,  7 km  south  of  Saumur  and  north-east  of  Montreuil-Bellay,  Maine- 
et-Loire,  and  numerous  specimens  in  the  Museum  de  Paleontologie,  Angers,  from  this  bed  and  adjacent  levels  in 
the  Tuffeau  Blanc  (Couffon  Collection  etc.)  variously  labelled  Saumoussay,  St.  Cyr-en-Bourg,  Saumur,  and 
elsewhere,  including  AM  57,  AM  59,  AM  116. 

There  are  numerous  specimens  from  Ponce,  Sarthe,  and  others  from  Bourre  in  the  Cher  Valley,  Loir-et-Cher, 
including  BMNH  C74803. 


Dimensions 


D 

Wb 

Wh 

Wb:  Wh 

U 

Ribs 

Lectotype 

130(100) 

40  (31) 

40  (31) 

1 

50  (38) 

13 

FSR,  C273 

67-3  (100) 

20-4  (30) 

25-0  (37) 

0-81 

23-9  (36) 

24 

MNHP  W7 

58-5  (100) 

21-0  (36) 

-(-) 

— 

22-8  (39) 

19 

MNHPW18 

61-0(100) 

24-0  (39) 

23-7  (39) 

101 

22-0  (36) 

22 

MNHP  X’ 

86-0  (100) 

29-0  (34) 

34-9  (41) 

0-83 

29-5  (34) 

~20 

MNHPW15 

81-0(100) 

32-0  (40) 

32-0  (40) 

1-0 

27-8  (34) 

19 

MNHPIc 

141-0(100) 

52-0  (37) 

49-5  (35) 

1-05 

55  (39) 

— 

Ic 

MNHP  6778 
MNHPW20 

133-0(100) 

41-5(29) 
45-0  (34) 

49- 5  (35) 

50- 0  (38) 

0-83 

0-9 

44-5  (33) 

18 

Ic 

162-0(100) 

55-0  (34) 

52-0  (32) 

1-05 

67  (41) 

— 

MNHP  W4  c 

137-0(100) 

60-0  (43) 

51-0(37) 

1-18 

- (-) 

15/16 

Ic 

39-5  (29) 

39-5  (29) 

1-0 

MNHP  W10 

109.0(100) 

47-5  (44) 

42-0  (39) 

M3 

37-0  (34) 

— 

33-5(31) 

38-8  (36) 

0-86 

MNHPW19 

175-0(100) 

74-8  (43) 

65-0  (37) 

1-15 

65-0  (37) 

15 

Description.  The  inner  whorls  of  our  smallest  specimens  show  coiling  to  have  been  moderately  evolute,  with 
compressed  whorls  and  a shallow  umbilicus.  At  about  10-15  mm  diameter,  there  are  27-32  ribs  per  whorl;  the 
density  decreases  with  increasing  size.  The  ribs  are  even,  bar-like,  prorsiradiate,  straight  and  clearly  demarcated 
from  the  flat  interspaces.  As  size  increases,  ribs  become  much  more  widely  spaced;  at  40-50  mm  diameter  there 
are  only  17-24  ribs  per  whorl.  They  are  of  variable  strength,  arise  from  weak  to  strong  umbilical  bullae  and  are 
narrow,  high  and  separated  by  wide,  flat  interspaces;  they  are  markedly  prorsiradiate  and  straight  to  concave  on 
the  flanks,  always  single,  with  no  intercalated  ribs.  At  the  ventrolateral  shoulder  they  bear  conical  to  feebly 
clavate  inner  ventrolateral  tubercles.  From  these  the  ribs  are  either  weakly  or  strongly  projected  forwards  to 
elongate  outer  ventrolateral  clavi.  A broadened  swelling  connects  these  in  turn  to  a sharp,  continuous  siphonal 
keel,  strengthened  into  sharp  high  clavi  at  the  peak  of  the  variably  angled  ventral  chevron  formed  by  the 
termination  of  the  ribs. 

This  type  of  ornament  may  extend  to  diameters  of  100  mm,  but  typically,  as  size  increases,  a series  of  changes 
in  ornament  occur,  more  or  less  independently  of  each  other.  The  umbilical  bullae  move  outwards  and  come  to 
occupy  a lower  flank  position,  whilst  the  ribs  are  differentiated  into  long  bullate  ones  and  (in  most  specimens) 
from  one  to  four  shorter  ribs,  restricted  to  the  outer  flank  and  venter  and  sometimes  lacking  ventrolateral 
tubercles.  The  inner  ventrolateral  tubercles  may  at  this  stage  develop  into  a distinctive  conical  horn  which 
supports,  on  the  outer  flank  of  its  base,  the  outer  ventrolateral  clavus;  some  specimens  present  a ventral  aspect  in 


EXPLANATION  OF  PLATE  63 

Figs.  1-12.  Collignoniceras  woo//gan'(Mantell).  1-4,  OUM  KZ  746;  11-12,  OUM  KZ  748,  from  the  St.  Cyr-en- 
Bourg  Fossil  Bed,  Champignonniere  les  Rochains,  south  of  Saumur  and  north-east  of  Montreuil-Bellay, 
Maine-et-Loire.  5-6,  MNHP  6778  (d’Orbigny  Collection),  Ponce,  Sarthe;  7-8,  OUM  KT  1160,  from  the 
Ojinaga  Formation  at  Cannonball  Hill,  northern  Chihuahua,  Mexico.  9,  Apertural  view  of  the  lectotype, 
BMNH  5682;  see  explanation  of  Plate  62  for  details.  10,  MNHP  Wl,  ‘Le  Mans,  Sarthe’  (from  Ponce?). 
Figures  1 -2  are  x 2;  the  remainder  are  x 1 . 


PLATE  63 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


f-ff- 


564 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  1.  Sutures  of  Collignoniceras  species.  A,  C.  woollgari  (Mantell),  from  BMNH  C74803;  B,  C.  carolinum 
(d’Orbigny),  from  the  Sorbonne  specimen  (de  Grossouvre  Collection);  c,  C.  papale  (d’Orbigny),  from  a 
Sorbonne  specimen  (de  Grossouvre  Collection).  Bar  scale  is  2 cm. 


which  siphonal  tubercles  greatly  outnumber  ventrolateral,  whilst  others  show  a more  or  less  equal  number;  no 
two  specimens  agree  in  details  of  ornament. 

Mature  specimens  show  two  broad  types  of  decoration,  but  again  no  two  specimens  agree  in  detail.  In  the  first 
group  the  umbilical  bullae  move  outwards  and  fuse  with  the  inner  ventrolateral  tubercles  to  form  a strong  to 
massive  horn  (if  broad)  or  flange  (if  narrow).  This  supports  a long,  low,  narrow  outer  ventrolateral  clavus,  and 
the  front  and  rear  of  the  horn  strengthens  into  a pair  of  ribs  which  loop  to  the  pair  of  siphonal  clavi 
corresponding  to  each  horn.  Some  specimens  may  develop  a low  siphonal  horn  at  this  stage  and  at  the  adult 
aperture  up  to  three  ventral  ribs  may  appear  between  the  primary  ribs,  although  in  other  specimens  these  may  be 
absent,  the  spaces  between  the  major  ribs  being  smooth.  The  second  type  is  a more  evolute  form,  retaining  long, 
straight,  distant  flank  ribs  with  bullae  of  variable  strength,  connected  by  weak  or  almost  effaced  ribs  to  strong 
conical  ventrolateral  horns  which  bear  the  outer  ventrolateral  clavus.  A low  siphonal  ridge  is  present  and  there 
are  pairs  of  clavi  corresponding  to  the  horns  as  well  as  additional  clavi  in  the  interspaces.  This  form  differs  most 
obviously  from  the  first  in  the  retention  of  bullae  and  in  being  somewhat  larger. 

The  suture  line  is  simple,  with  a massive,  slightly  incised,  asymmetrically  bifid  E/L,  narrow  L and  narrow,  bifid 
L/U2. 


EXPLANATION  OF  PLATE  64 

Figs.  1-3.  Collignoniceras  woollgari  (Mantell).  The  lectotype  of  Acanthoceras  schlueterianum  (Laube  and 
Bruder),  from  the  Turonian  of  the  White  Mountain  near  Prague,  Czechoslovakia.  Pictures  supplied  through 
the  courtesy  of  Dr.  V.  Housa  (Prague). 


PLATE  64 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


566 


PALAEONTOLOGY,  VOLUME  23 


Discussion.  The  above  description  is  based  upon  the  available  English  material,  the  large  suite  of 
specimens  from  Touraine  and  a few  Czechoslovakian  specimens  before  us.  It  must  be  stressed  that  no 
two  specimens  are  alike  and  that  description  is  inevitably  generalized.  Mantell’s  original  figures  of 
Ammonites  woollgari  give  a clear  and  accurate  representation  of  the  juvenile  form,  but  only  suggest 
the  very  different  adult  form  in  general  terms,  better  shown  in  Sowerby’s  (1828,  p.  165;  pi.  587,  fig.  1) 
beautiful  watercolour  and  Sharpe’s  (1855,  p.  27;  pi.  11,  figs,  la-b)  slightly  inaccurate  reconstruction. 

The  lectotype  is,  in  fact,  a moderately  distorted  composite  internal  mould  only  130  mm  in 
diameter,  as  can  be  seen  from  our  photographs  (PI.  62,  figs.  1-2;  PI.  63,  fig.  9),  showing  no  trace  of 
sutures  or  any  indications  of  how  much  is  body  chamber.  In  terms  of  the  description  given  above,  it 
falls  into  the  first  group  of  specimens.  It  is  distinctive  in  the  small  size  at  which  the  massive  horns  are 
developed  and  the  brevity  of  the  stage  with  intercalated  ribs. 


text-fig.  2.  Collignoniceras  woollgari  (Mantell)  BMNH  88989a,  a crushed  specimen 
from  the  Turonian  of  the  White  Mountain,  near  Prague,  Czechoslovakia. 


text-fig.  3.  Collignoniceras  woollgari  (Mantell)  a,  b,  MNHP  W14,  6778  (d’Orbigny  Collection),  from  Ponce, 
Sarthe.  a.  tuffeau  specimen  agreeing  closely  with  the  type.  Reduced  x 0-5  approx,  c.  d.  MNHP  1946-19,  from  St. 
Maure  de  Touraine.  A hypernodose  adult  of  the  first  type.  Reduced  x 0-4  approx. 


568 


PALAEONTOLOGY,  VOLUME  23 


At  the  beginning  of  the  outer  whorl  the  ribs  bear  strong  umbilical  bullae,  strong  conical  inner 
ventrolateral  and  long,  low,  clavate  outer  ventrolateral  tubercles  and  a strong  elongate  siphonal 
clavus.  Between  these  long  primary  ribs  are  one  or  two  shorter  intercalated  ribs  which  extend  across 
the  venter  and  bear  small  siphonal  clavi.  By  90  mm  diameter  these  are  lost  and  the  ornament  consists 
of  an  umbilical  bulla  which  moves  out  progressively  to  occupy  a mid-lateral  position,  linked  by  a 
broad  rib  to  a massive  inner  ventrolateral  horn  which  bears,  at  its  base,  the  outer  ventrolateral  clavus. 
From  this  clavus  two  poorly  defined,  low,  rounded  ribs  link  to  two  ventral  clavi. 

The  best-preserved  horn  on  the  lectotype  is  at  120  mm  diameter,  and  here  the  bulla  on  the  flank  and 
the  inner  ventrolateral  horn  have  merged  into  a massive  horn  bearing  a much  weakened  outer 
ventrolateral  clavus  and  subdued  weakened  ribs. 

D’Orbigny  (1841,  p.  352,  pi.  108,  figs.  1-3)  figured  under  the  name  A.  woollgari  a distinctive  form 
which  he  subsequently  (1850)  named  A.  vielbancii;  it  is  redescribed  below  as  a junior  subjective 
synonym  of  Lecointriceras  fleuriausianum  (d’Orbigny).  D’Orbigny  also  described  in  Paleontologie 
Franqaise  a related  form,  A.  carolinus  (1841,  p.  310,  pi.  91,  figs.  5-6),  which  he  subsequently  (1850) 
regarded  as  a synonym  of  A.  woollgari , a view  followed  by  most  later  authors.  Sharpe  (1855,  p.  27, 
pi.  11,  figs,  la-b,  2a-b)  clearly  recognized  the  differences  between  young  woollgari  and  carolinum 
(\  • ■ the  French  shell  has  twice  as  many  ribs,  is  less  compressed,  and  has  the  keel  more  completely 
separated  from  the  ribs  by  two  regular  channels,  than  in  our  species’),  and,  as  we  describe  below,  the 
two  are  indeed  specifically  distinct. 

Fritsch  (1872)  provided  a very  clear  discussion  of  Mantell’s  species,  and  recognized  three  variants; 
his  descriptions  are  loosely  translated  as  follows: 

(a)  Typical  form,  which  agrees  exactly  with  the  illustrations  of  Mantell  and  Sharpe.  It  has  very 
strong  tubercles  on  the  siphonal  side  (pi.  4,  figs.  1,  2). 

(b)  Form  with  slender  ribs  and  weaker  tubercles  (pi.  3,  fig.  2). 

( c ) More  involute  form  with  an  inverse  egg-shaped  mouth  opening.  There  are  tubercles  close  to  the 
umbilical  seam,  which  remain  there  for  a long  period,  and  are  stronger  and  more  widely  separated 
than  in  the  typical  form;  there  are  only  six,  even  on  the  inner  whorl  (pi.  3,  fig.  1). 

He  also  described  a variety  lupulina  from  Mecholup  [Michelob]  near  Saatz,  close  to  Prague  (1872, 
p.  31,  pi.  2,  figs.  1,  2;  pi.  15,  fig.  6),  which  was  said  to  be  very  similar  to  woollgari  when  young,  but 
when  old,  has  a different  venter,  large  sparse  tubercles  and  an  almost  square  cross-section.  It  is, 
in  fact,  a Mammites  nodosoides  (Schliiter). 

Schliiter  ( 1 872)  figured  a similar  range  of  variants;  his  pi.  9,  figs.  1 -3  correspond  to  Fritsch’s  form  c 
and  his  pi.  9,  figs.  4-5  to  the  typical  form.  His  variety  (pi.  12,  figs.  5-6)  is,  as  he  suggested,  close  to  the 
papale  group  in  many  respects  and  it  could  well  be  referred  to  as  Collignoniceras  aflf.  canthus  (Sornay). 

Laube  and  Bruder  (1887)  reviewed  a similar  range  of  central  European  specimens  but  referred 
Fritsch’s  typical  form  (var.  a)  to  a new  species,  Acanthoceras  schlueterianum;  they  regarded  the 
involute  form  (var.  c)  as  typical  C.  woollgari  and  var.  lupulina  as  a Mammites,  which  they  renamed 
Mammites  michelobensis.  Petrascheck  (1902)  followed  Laube  and  Bruder  and  described  forms  he 
called  woollgari,  schlueterianum,  and  aff.  woollgari. 

From  our  study  of  the  type  material  and  the  Touraine  populations,  it  is  quite  clear  that  no  two 
adult  Collignoniceras  of  these  types  are  the  same.  The  lectotype  of  C.  woollgari,  showing  as  it  does  an 
early  loss  of  umbilical  bullae,  which  move  out  to  mid  flank,  fuse  into  ventrolateral  horns,  with  much 
elongated  outer  ventrolateral  clavi  and  subdued  ribs  looping  to  low  siphonal  clavi  is  clearly  of  the 
same  general  morphology  as  Fritsch’s  typical  form  (e.g.  1872,  pi.  4,  figs.  1-2)  and  the  lectotype  (here 
designated)  of  Acanthoceras  schlueterianum  (Laube  and  Bruder  1887,  pi.  29,  figs.  2a-b)  (PI.  64).  It 
differs,  however,  in  showing  a decline  in  ventral  ribs  and  clavi  at  only  90  mm  diameter,  whereas  the 
Czechoslovakian  examples  retain  umbilical  bullae  and  intercalated  ribs  (particularly  on  the  venter)  to 
a much  greater  size  and  in  consequence  have  a longer  middle  growth  stage  with  umbilical  bullae, 
conical  inner  ventrolateral  and  outer  ventrolateral  and  siphonal  clavi,  like  the  specimen  illustrated 
here  (text-fig.  4 c-d),  Fritsch’s  pi.  14,  fig.  6 and  Laube  and  Bruder’s  smaller  paralectotype  (1887, 
pi.  29,  fig.  3).  This  stage  is  virtually  suppressed  in  the  lectotype  of  C.  woollgari,  which  in  these  respects 


text-fig.  4.  Collignoniceras  woollgari  (Mantell)  a,  b.  MNHP  W22,  6778  (d’Orbigny  Collection),  from  Ponce, 
Sarthe.  An  adult  of  the  second  type,  retaining  long  ribs  and  moderately  evolute  coiling.  Reduced  x 0-4  approx, 
c,  D.  BMNH  88988b,  from  the  Turonian  of  the  White  Mountain,  near  Prague.  Reduced  x0-5  approx. 


570 


PALAEONTOLOGY,  VOLUME  23 


is  atypical.  Other  specimens  show  that  the  intercalated  ventral  ribs  are  accompanied  by  weak  flank 
ribs  in  middle  growth  but  that  there  is  great  variation  at  this  stage.  The  Touraine  populations,  which 
yield  specimens  that  both  match  the  lectotype  of  C.  woollgari  (text-fig.  3 a-b)  and  show  every 
gradation  to  the  other  forms  (PI.  66,  figs.  1-3;  text-fig.  3 c-d)  with  strong  intercalated  ribs  and 
tubercles,  show  that  C.  woollgari  and  C.  schlueterianum  should  be  treated  as  synonyms.  Indeed,  a 
specimen  from  Fritsch’s  own  collection,  now  in  the  British  Museum  (Natural  History)  (no.  88989a) 
and  labelled  in  his  own  hand  ‘Weisser  Berg’,  the  type  locality  of  C.  schlueterianum  (text-fig.  2), 
exhibits  the  fusion  of  umbilical  bullae  with  strong  horns  seen  in  the  lectotype  of  woollgari  but  with 
more  persistent  intercalated  ribs  on  the  venter  of  the  last  whorl.  The  specimen  is,  furthermore,  adult 
at  only  150  mm,  showing  a rapid  decline  in  ornament  and  loss  of  horns  on  the  outer  whorl. 

In  Germany  (?),  Czechoslovakia  and  Touraine  (but  not  England  where  only  one  adult  is  known) 
this  hypernodose,  horned  form,  enormously  variable  in  its  adult  ornament,  is  accompanied  by  the 
evolute,  square-whorled  forms  which  correspond  to  Fritsch’s  form  C,  to  Laube  and  Bruder’s 
‘typical  form’  and  Petrascheck’s  A.  woollgari  + A.  cfr.  woollgari.  Inner  whorls  of  this  type  are 
inseparable  from  typical  juvenile  English  C.  woollgari,  but  again  the  variable  adult  whorls  are  quite 
distinctive,  as  Fritsch  described,  and  as  outlined  above  in  our  description;  we  conclude  that  these  are 
probably  sexual  dimorphs. 

C.  woollgari  var.  mexicana  (Bose)  (1928,  p.  262,  pi.  11,  figs.  11,  12)  was  originally  described  on  the 
basis  of  a single,  crushed  specimen  from  the  Turonian  Ojinaga  Formation  equivalent,  near  Jimenez, 
Coahuila,  Mexico,  reillustrated  here  as  Plate  65,  figs.  1-3.  Powell  (1963)  has  redescribed  this  form  (as 
Selwynoceras  mexicanum ) and  discussed  the  intraspecific  variation  on  the  basis  of  large  collections  of 
fragmentary  material.  From  large  additional  collections  from  the  same  area  (OUM  KT  1 160-1183, 
1200-1222,  1264-1313)  and  Chispa  Summit,  Jeff  Davis  County,  Texas  and  specimens  in  the  Adkins 
Collection  (preserved  in  the  Texas  Memorial  Museum)  we  conclude  that  it  too  is  a synonym  of 
C.  woollgari.  Juveniles,  as  Powell  himself  noted  (op.  cit.,  p.  1225),  include  individuals  which  cannot 
be  separated  from  the  English  C.  woollgari  (PI.  63,  figs.  7-8),  in  addition  to  those  which  are  more 
compressed,  finely  and  densely  ribbed. 

Powell  (1963,  pi.  168,  fig.  4)  has  figured  a specimen  in  middle  growth,  showing  the  irregularly 
ribbed  stage  with  development  of  inner  ventrolateral  horns  as  seen  in  Bohemian  and  Touraine 
specimens  and  we  have  other  slender  fragments  which  match  Petrascheck’s  (1902,  pi.  10,  figs.  3a-b) 
juvenile  A.  schliiterianum.  Larger  fragments  show  a wide  range  of  variation,  from  robust  fragments 
having  essentially  equal  numbers  of  inner  and  outer  ventrolateral  and  siphonal  tubercles  to  those 
with  multiple  ventral  tuberculation.  Adult  body  chambers  show  clear  dimorphism,  as  in  European 
material,  the  one  form  with  flanges  or  flared  horns  produced  by  amalgamation  of  umbilical  and 
ventrolateral  tubercles,  the  other  more  quadrate,  retaining  to  maturity  umbilical  bullae  and  distant 
ribs  of  variable  strength.  As  can  be  seen  from  our  and  Powell’s  figures,  distinction  on  the  basis  of  the 
nature  of  the  less  complex  suture,  the  finer  ribbed  juveniles  and  the  coarse  ornament  of  adults,  by 
which  Powell  separated  it  from  C.  schlueterianum,  cannot  be  upheld  in  the  light  of  the  variation  seen 
in  European  specimens  (not  known  to  Powell);  there  is  a clear  overlap.  We  note  the  relatively 
frequent  occurrence  of  individuals  with  flares  and  a compressed  whorl,  rarely  seen  in  Europe, 
suggesting  the  Texas/Mexico  material  belongs  to  a local  population  more  variable  than  their  old 
world  contemporaries. 


EXPLANATION  OF  PLATE  65 

Figs.  1-8.  Collignoniceras  woollgari  (Mantell).  1-3,  the  holotype  of  Prionotropis  woollgari  (Mantell)  var. 
mexicana  Bose,  from  near  Jimenez,  Coahuila,  Mexico.  University  of  California,  Berkeley,  Collections.  4-6, 
BMNH  4863a,  from  the  Middle  Chalk  near  Lewes,  Sussex.  7-8,  a juvenile  U.S.  Western  Interior  specimen  in 
the  U.S.  Geological  Survey  Collections,  Denver,  from  USGS  Mesozoic  locality  21792,  the  mid-Turonian 
Carlile  Shale  of  the  Black  Hills. 


PLATE  65 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


572 


PALAEONTOLOGY,  VOLUME  23 


The  relationship  of  European  specimens  to  the  widely  documented  U.S.  Western  Interior  material 
referred  to  C.  woollgari  has  been  complicated  by  the  relatively  few  illustrations  of  English  juveniles. 
Adults  such  as  Meek’s  specimen  (1876,  pi.  7,  fig.  lg)  from  the  Black  Hills,  Dakota,  would  certainly 
fall  within  the  concept  of  C.  woollgari  outlined  here,  although  differing  from  the  lectotype  most 
obviously  in  the  retention  of  umbilical  bullae  to  a greater  diameter.  Dr.  W.  A.  Cobban  (Denver)  has 
also  shown  us  medium-sized  specimens  in  which  all  ribs  are  long  and  the  ventrolateral  and  siphonal 
clavi  are  equal  in  number,  a feature  uncommon  in  European  material.  American  juveniles,  described 
by  Haas  (1946)  and  Matsumoto  (1965)  amongst  others,  show  a much  wider  range  of  variation  than 
European  material.  This  may  be  merely  a consequence  of  the  small  number  of  juveniles  known  from 
Europe:  indeed,  the  latter  fall  closest  to  Matsumoto’s  group  E,  one  of  the  commonest  forms  in  the 
Western  Interior.  Nevertheless,  there  is  a clear  overlap  with  European  C.  woollgari.  The  presence  of 
similar  individuals  would  also  seem  to  preclude  subspecific  separation  and  we  regard  them  as 
conspecific,  but  with  a different  population  structure.  Specific  differentiation  of  the  American  fauna 
from  their  European  contemporaries  occurred  later,  with  the  evolution  of  the  early  members  of  the 
Prionocyclus  hyatti  group. 

W.  A.  Cobban  (in  lift.)  has  suggested  to  us  that  forms  with  more  siphonal  than  ventral  nodes  pre- 
date those  in  which  the  numbers  are  equal  in  the  U.S.  Western  Interior,  but,  as  we  do  not  know  the 
precise  horizon  of  the  holotype  of  woollgari  in  relation  to  these,  we  prefer  to  unite  them  here,  leaving 
revision  of  these  faunas  to  Dr.  Cobban. 

According  to  Matsumoto  (1959,  p.  107;  1965,  p.  16,  pi.  3,  figs.  3-4)  C.  woollgari bakeri  Anderson  is 
a subgroup  of  C.  woollgari  that  characterizes  the  north  Pacific  region.  All  described  specimens  are 
small,  compressed,  evolute  Subprionocylus- like  densely  ribbed  shells,  close  to  subgroup  D of 
C.  woollgari  of  Matsumoto  (1965)  from  the  U.S.  western  Interior,  but  more  evolute  and  with  less 
prorsiradiate  ribs.  These  differences  probably  do  not  merit  separation,  but  without  more  and  adult 
specimens  further  comment  is  inadvisable. 

C.  woollgari  is  easily  separated  from  the  remaining  species  of  the  genus.  C.  carolinum  (d’Orbigny) 
(p.  574)  is  usually  more  densely  ribbed  and  even  in  sparsely  ribbed  juveniles  (PI.  68,  fig.  1 1)  the  ribs  are 
low  and  subdued  rather  than  bar-like.  Adults  are  quite  distinct;  C.  carolinum  reaches  maturity  at  only 
100-120  mm  diameter,  never  develops  the  coarse  umbilical  bullae,  ribs,  and  horns  of  woollgari,  nor 
the  complex  looped  ventral  ornament.  Instead,  it  remains  compressed  and  flat  sided,  with  weak  ribs 
and  tubercles  and  a persistent,  crenulated  siphonal  ridge.  C.  canthus  (Sornay)  (p.  582)  has  coarsely 
and  sparsely  ribbed  and  tuberculate  inner  whorls  but  a virtually  smooth  body  chamber  with  only 
faint  ribs  and  many  tiny  siphonal  tubercles.  C.  turoniense  (Sornay)  (p.  584)  has  similarly  coarsely 
ornamented  early  whorls,  is  adult  at  a much  smaller  size  with  more  massive  whorls,  coarse  sparse 
bullae,  weak  ribs  and  ventrolateral  horns  and  the  inner  ventrolateral  tubercles  disappear  at  an  early 
stage. 

There  is  a closer  resemblance  to  C.  papale  (d’Orbigny)  (p.  578)  but  here  juveniles  have  fewer, 
coarser  ribs  with  strong  bullae  displaced  well  out  from  the  umbilical  shoulder,  with  much  more 
prominent  inner  ventrolateral  tubercles.  In  middle  growth  C.  papale  lacks  the  prominent 
ventrolateral  horns  of  many  C.  woollgari  and  the  inner  and  outer  ventrolateral  tubercles  merge  into  a 
pinched  clavus,  retained  to  much  greater  diameters  in  C.  woollgari.  Other  differences  are  noted  on 
p.  582. 

C.  boreale  (p.  586)  is  a genuinely  small  form,  showing  adult  features  at  only  100  mm  diameter  in  the 
holotype.  It  has  narrow,  distant  ribs  and  retains  umbilical  bullae  to  the  end  of  the  phragmocone, 
showing  early  development  of  flared  ventrolateral  flanges  and  traces  of  looped  ventral  ribs. 


EXPLANATION  OF  PLATE  66 

Figs.  1-3.  Collignoniceras  woollgari  (Mantell).  Adult  phragmocone  showing  intercalation  of  flank  and  ventral 
ribbing,  multiple  ventral  tuberculation  and  early  stages  of  horn  development.  MNHP  W10,  from  either 
Ponce  (Sarthe)  or  Bourre  (Loir-et-Cher). 


PLATE  66 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


574 


PALAEONTOLOGY,  VOLUME  23 


Occurrence.  Few  C.  woollgari  from  England  are  well  dated.  Through  the  courtesy  of  the  Director  of  the  Institute 
of  Geological  Sciences  and  Mr.  C.  J.  Wood  we  have  been  able  to  examine  the  precisely  positioned  material  from 
the  Leatherhead  (Fetcham  Mill),  Surrey,  Borehole  (Gray  1965).  Here  C.  cf.  woollgari  occurs  at  a depth  of 
570'  6”  (GSM.WN  1979-80, 1982-3),  73'  1"  (22-28  m)  above  the  base  of  the  Melbourn  Rock  and  17'  6"  (5-33  m) 
above  a specimen  of  IMytiloides  hercynicus ; at  535'  10"  (GSM.WN  1942),  12'  (3-66  m)  above  the  level  of  large 
Inoceramus  of  inequivalvis  type,  and  at  518'  9"  (GSM.WN  1900,  1901),  26'  9"(8-15  m)  below  specimens  of 
Mytiloides  sp.  and  /.  cf.  apicalis  (inoceramids  determined  by  Mr.  P.  Woodroof).  This  range,  through  51'  9" 
(15-8  m)  of  section,  includes  the  top  of  the  Inoceramus  labiatus/Orbirhynchia  cuvieri  and  the  lower  part  of  the 
Terebratulina  lata  Zones.  Other  English  specimens  have  been  recorded  from  both  labiatus  and  lata  Zones. 
Specimens  from  Sussex,  the  type  area,  come  mostly  from  the  Lewes  region.  One  specimen  (BMNH  C30394)  is 
said  to  be  from  Mount  Caburn;  unfortunately  the  classic  pit  here  extends  from  the  Melbourn  Rock  to  basal 
Upper  Chalk  {labiatus -planus  Zones). 

Specimens  from  the  upper  part  of  the  lata  Zone  of  Surrey  (e.g.  WW  14792-4,  16682),  and  OUM  K 10273, 
K 10275-6  from  no  more  than  5 m below  the  top  of  the  Chalk  Rock  at  Fognam,  Berkshire,  indicate  the  upper 
limit  of  its  relatively  long  range.  This  is  confirmed  by  occurrences  in  Sarthe  and  Touraine  through  the  middle  and 
upper  part  of  the  Tuffeau  Blanc,  in  the  St.  Cyr-en-Bourg  Fossil  Bed,  Bourre  and  Ponce  faunas.  In  the  United 
States  the  species  occurs  rarely  in  the  top  of  Cobban  and  Scott’s  (1972)  Mammites  nodosoides  Zone  (Cobban  in 
lilt.)  and  overlaps  with  the  succeeding  Prionocyclus  hyatti  (Powell,  1963). 

Elsewhere  the  species  is  known  to  occur  widely  in  Europe,  the  U.S.S.R.  west  to  Transcaspia,  Japan,  California 
and  Oregon,  Texas,  Mexico,  the  U.S.  Western  Interior  and  northern  Australia. 

Collignoniceras  carolinum  (d’Orbigny) 

Plate  68,  figs.  1-11;  Plate  76,  figs.  1-2;  text-figs.  1b,  5 

1841  Ammonites  Carolinus  d’Orbigny,  p.  310,  pi.  91,  figs.  5-6. 

1850  Ammonites  Woolgarii  Mantell;  d’Orbigny,  p.  189  (pars). 

1860  Ammonites  Carolinus  d’Orbigny;  Pictet  and  Campiche,  p.  316. 

1872  Ammonites  carolinus  d’Orbigny;  Schliiter,  p.  27,  pi.  9,  fig.  6. 

1881  Ammonites  Carolinus  d’Orbigny;  Windmoller,  p.  33. 

71887  Acanthoceras  Carolinum  d’Orbigny;  Laube  and  Bruder,  p.  232,  pi.  27,  fig.  1. 

1902  Prionotropis  carolinus  d’Orbigny;  Petrascheck,  p.  152. 

71912  Prionotropis  woolgari  var.  Carolinus  (d’Orbigny);  Arkhanguelsky,  p.  72,  pi.  3,  figs.  20-22  ( fide 
Arkhanguelsky,  1916). 

1925  Prionotropis  Carolina  (d’Orbigny);  Diener,  p.  156  (pars). 

1977  Collignoniceras  (Collignoniceras)  carolinum  (d’Orbigny);  Hancock,  Kennedy  and  Wright, 
p.  156. 

Types.  D’Orbigny’s  original  account  of  this  species  is  as  follows:  ‘Je  l’ai  recueillie  en  place  aux  Martrous,  pres 
de  Rochefort  (Charente-Inferieure),  dans  la  craie  que  je  rapporte  aux  gres  verts  superieurs  ou  aux  craies 
chloritees.  Elle  y est  rare  a l’etat  de  moule.  M.  d’Archiac  l’a  aussi  rencontree  a Sainte-Maure  (Indre-et-Loire), 
dans  le  meme  couche.’  By  1850  d’Orbigny  had  concluded  that  carolinus  was  a synonym  of  woollgari  ( Prodrome , 
p.  189),  and  in  consequence  no  specimens  are  represented  in  his  collections  under  the  name  carolinus.  Under 
Ammonites  woollgari,  however,  there  is  a specimen  from  Martrous  with  the  label  6778a  which  is  clearly  the  basis 
of  the  original  figure  (PI.  68,  figs.  4-8),  and  this  is  here  designated  lectotype  of  the  species. 

Other  specimens  studied.  OUM  KZ  747,  from  the  St.  Cyr-en-Bourg  Fossil  Bed,  Champignonniere  les  Rochains, 
7 km  south  of  Saumur  and  north-east  of  Montreuil-Bellay,  Maine-et-Loire.  An  unregistered  specimen  in  de 
Grossouvre’s  collection  (Sorbonne,  Paris)  from  either  Ponce  (Sarthe)  or  Bourre  (Loir  et  Cher).  MNHP  W8, 
from  an  unknown  locality  in  the  Tuffeau.  WW  14791  from  the  Terebratulina  lata  Zone,  Mickleham  Bypass, 
Surrey. 


EXPLANATION  OF  PLATE  67 

Figs.  1-3.  Collignoniceras  woollgari  (Mantell).  Adult  phragmocone  of  sparsely  and  robustly  ribbed  variant 
with  equal  numbers  of  umbilical,  ventrolateral  and  siphonal  tubercles.  MNHP  W2.  1904-32.  ‘Le  Mans, 
Sarthe’. 


PLATE  67 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


576 


PALAEONTOLOGY,  VOLUME  23 


Dimensions 


D 


Wb 


Wh 


Wb:  Wh  U 


Lectotype 
MNHP  6778a 
Sorbonne  spec. 


46-0(100)  14-0(30)  15-0(33)  0-93  16-7  (36) 

108-5  (100)  28-2(26)  37-5(35)  0-75  34-3  (32) 


Description.  The  lectotype  from  Martrous  (Charente-Maritime)  is  a fragment  with  juvenile  body  chamber 
preserved  in  calcarenite  typical  of  the  Calcaires  a Cephalopodes  of  the  Rochefort  area.  Coiling  is  relatively 
evolute,  with  a broad,  shallow  umbilicus  (36%  of  the  diameter).  The  umbilical  wall  is  low  and  rounded.  The 
whorl  section  is  compressed  (whorl  breadth  to  height  ratio  is  approximately  0-93),  with  flattened,  convergent 
sides,  the  maximum  breadth  close  to  the  umbilical  shoulder  and  the  venter  fastigiate.  Ornament  consists  of 
strong,  dense,  narrow  ribs  (nineteen  on  last  half-whorl),  arising  at  the  umbilical  shoulder  without  clear  bullae 
after  the  first  two  visible  ribs.  They  are  straight  or  slightly  flexed  and  prorsiradiate  on  the  inner  flank,  curving 
strongly  forwards  across  the  ventrolateral  shoulders  and  venter.  Single,  shorter  intercalated  ribs  occur 
commonly  on  the  early  part  of  the  specimen  but  there  are  only  two  in  the  last  half- whorl.  The  ribs  are 
strengthened  into  distinct  if  small  inner  ventrolateral  tubercles  at  the  beginning  of  the  body  chamber,  but  these 
are  lost  beyond  a diameter  of  about  34  mm.  There  are  well-marked  outer  ventrolateral  clavi,  connected  by 
forwards-directed  weak  ribs  to  elongate  siphonal  clavi  borne  on  a low,  rounded  keel.  Other  specimens  show  both 
denser  and  sparser  ribbing  of  the  same  style,  as  in  other  Collignoniceras  juveniles  (PI.  68,  figs.  10,  1 1). 

Body  chambers  show  the  species  to  have  been  adult  at  small  diameters  (100-120  mm).  The  adult  whorls  are 
compressed  (whorl  breadth  to  height  ratio  as  little  as  0-75)  with  gently  inflated  inner,  and  flattened  outer  flanks, 
with  a fastigiate  venter.  Ornament  consists  of  numerous  (about  thirty)  rather  low,  rounded,  prorsiradiate  ribs, 
arising  at  the  umbilical  shoulder  without  bullae  and  flexed  strongly  forwards,  concave  on  the  outer  flank  and 
ventrolateral  shoulder,  where  they  bear  blunt,  clavate  tubercles.  The  ribs  are  narrow  as  they  sweep  forwards 
from  these  to  long  siphonal  clavi.  Rarely  ribs  branch  from  the  umbilical  seam  or  are  intercalated,  so  that  there 
are  more  siphonal  clavi  than  long  ribs. 

The  sutures  are  indifferently  exposed  (text-fig.  1 b),  but  are  typically  collignoniceratid,  with  broad,  simple,  bifid 
elements. 

Discussion.  D’Orbigny’s  figure  is  partly  idealized:  in  addition  the  figure  lacks  the  abrupt  start  of  the 
ribs  at  the  umbilical  shoulder,  shows  too  many  short  ribs  and  makes  the  species  appear  too  inflated 
(text-fig.  5).  Pictet  and  Campiche  (1860,  p.  316)  and  de  Grossouvre  (1894,  p.  75)  regarded  this  species 
as  a juvenile  C.  woollgari,  but  Sharpe  (1855,  p.  27)  had  already  noted  that  ‘the  French  shell  has  twice 


Figs.  1-11.  Collignoniceras  carolinum  (d’Orbigny).  1-3,  SP,  de  Grossouvre  Collection,  probably  from  Bourre 
(Loir-et-Cher).  4-8,  the  lectotype,  MNHP  6778a,  from  the  Calcaire  a Cephalopodes  of  Martrous,  near 
Rochefort  (Charente-Maritime).  9-10,  OUM  KZ  747,  from  the  St.  Cyr-en-Bourg  Fossil  Bed,  Champignon- 
niere  les  Rochains,  south  of  Saumur  and  east  of  Montreuil-Bellay  (Maine-et-Loire).  1 1,  MNHP,  from  an 
unknown  locality  in  the  Tuffeau  Blanc  de  Touraine. 


text-fig.  5.  Collignoniceras  carolinum  (d’Orbigny). 
Copies  of  d’Orbigny’s  original  figures  (1841,  pi.  91, 
figs.  5-6). 


EXPLANATION  OF  PLATE  68 


PLATE  68 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


578 


PALAEONTOLOGY,  VOLUME  23 


as  many  ribs,  is  less  compressed,  and  has  the  keel  more  completely  separated  from  the  ribs  by  two 
regular  channels’.  Schliiter  (1872,  p.  27)  maintained  the  species,  as  did  Laube  and  Bruder  (1887, 
p.  232),  although  their  specimen  is  only  doubtfully  referable  to  it.  Meek  (1876,  p.  457)  regarded 
d’Orbigny’s  Ammonites  bravaisianus  as  the  juvenile  of  carolinum,  which  he  in  turn  treated  as  a 
synonym  of  woollgari. 

In  the  last  50  years  the  name  has  dropped  out  of  currency.  The  most  recent  reference  was  by 
Matsumoto  (1971,  p.  131)  who  upheld  the  view  that  it  was  possibly  an  immature  example  of  C. 
woollgari  in  which  the  appearance  of  strong  distant  ribs  was  delayed,  in  this  respect  being 
intermediate  between  C.  woollgari  woollgari  and  C.  woollgari  bakeri. 

C.  carolinum  is  in  fact  quite  distinct  from  C.  woollgari.  As  early  authors  noted,  the  type  of  the 
species  is  consistently  more  finely  and  densely  ribbed  than  European  C.  woollgari  and  at  comparable 
diameters  the  ribbing  is  much  more  subdued  and  the  ventral  tuberculation  finer.  Other  examples 
before  us  show  much  sparser  ribbing  (PI.  68,  fig.  11),  but  even  here  the  ribbing  is  more  subdued.  When 
adult  the  species  are  very  distinct;  C.  carolinum  reaches  maturity  at  only  100-120  mm  and  never 
develops  the  coarse  umbilical  bullae  and  ribs,  the  massive  ventrolateral  horns  or  the  complex  looped 
ventral  ribbing  and  tubercles  of  C.  woollgari. 

The  delicately  ribbed  inner  whorls  immediately  distinguish  the  species  from  the  grossly  tuberculate 
young  of  C.  canthus,  C.  turoniense  and  C.  papale.  Adult  C.  canthus  are  broader  whorled  and  retain 
massive  bullae  and  ribs,  whilst  C.  papale  has  strong  ribs  with  conspicuous  looping  as  well  as  being 
more  inflated.  The  feebly  ornamented  body  chamber  of  C.  turoniense  is  superficially  similar,  but  is 
much  broader,  virtually  lacks  ribs  but  has  a row  of  small  siphonal  tubercles. 

C.  boreale,  although  adult  at  a similarly  small  diameter,  has  much  coarser  ribbing  when  young,  and 
develops  distant  coarse  flared  ribs  when  adult. 

The  confusion  of  C.  carolinum  with  C.  woollgari  stems  from  the  similarity  of  the  former  to  finely 
ribbed  forms  of  the  latter  known  from  Japan  and  the  United  States.  These  have  been  described  by 
Haas  (1946)  as  Prionotropis  woollgari  vars.  regularis  and  tenuicostata,  and  by  Matsumoto  (1965)  as 
his  Group  B of  C.  woollgari.  These  finely  ribbed  forms  are  distinguished  from  the  type  of  C.  carolinum 
in  always  developing  relatively  coarse  ribs  at  a diameter  of  20  mm  or  less  and  by  ribs  that  are  sharp 
rather  than  subdued,  straight  rather  than  flexuous. 

Occurrence.  This  is  a rare  species.  Apart  from  the  Touraine  records  above,  it  is  known  in  France  from  the 
environs  of  La  Rochelle  in  Charente;  in  England  from  the  Terebratulina  lata  Zone  of  Surrey;  in  north  Germany, 
Bohemia  and  Turkestan. 


Collignoniceras  papale  (d’Orbigny) 

Plate  69,  figs.  1,  2;  Plate  70,  figs.  3-5;  text-figs,  lc,  6-7 

1841  Ammonites  Papalis  d’Orbigny,  p.  354,  pi.  109,  figs.  1-3. 

1850  Ammonites  papalis  d’Orbigny,  p.  189. 

1887  Acanthoceras  papaliforme  Laube  and  Bruder,  p.  237,  pi.  27,  figs.  3-4. 

1925  Prionotropis  papalis  d’Orbigny;  Diener,  p.  156. 

1925  Prionotropis  papaliformis  Laube  and  Bruder;  Diener,  p.  156. 

1977  Collignoniceras  ( Selwynoceras ) aff.  papale  (d’Orbigny);  Hancock,  Kennedy  and  Wright, 
p.  156. 

1977  Collignoniceras  ( Selwynoceras ) gr.  papale  (d’Orbigny);  Hancock,  Kennedy  and  Wright,  p.  156. 

Holotype.  By  monotypy  the  specimen  in  the  Requien  Collection  (Musee  d’ Avignon),  presumed  to  come  from  the 
'craie  tuffeau  ou  chloritee  du  departement  de  Vaucluse’  (d’Orbigny  1 841 , p.  356).  We  have  not  seen  the  holotype, 
but  d’Orbigny’s  figure  (text-fig.  6)  is  little  more  than  two-thirds  natural  size. 

Specimens  studied.  There  is  a series  of  specimens  in  the  Museum  d’Histoire  Naturelle,  Paris;  five  recorded  in  the 
d’Orbigny  Collection  as  coming  from  Montrichard  (Loir-et-Cher),  reg.  no.  6780;  MNHP  W.9,  unlabelled  but 
probably  from  Bourre;  MNHP  ‘3’,  from  Montrichard;  MNHP  ‘A’  B’  D’-‘E’  from  Bourre.  MNHP  ‘F’ 
unlocalized  but  from  the  Tuffeau  de  Touraine. 


KENNEDY  ET  AL.\  COLLIGNONICERATID  AMMONITES 


579 


There  are  several  unregistered  specimens  in  the  de  Grossouvre  Collection,  housed  in  the  Sorbonne,  from  either 
Bourre  or  Ponce;  a specimen  labelled  Bourre  showing  the  inner  whorls;  and  a small  body  chamber,  also 
unregistered,  is  labelled  Bourre. 

OUM  KZ  738  and  745  are  from  the  St.  Cyr-en-Bourg  Fossil  Bed,  Champignonniere  les  Rochains,  7 km  south 


of  Saumur  and  north-east  of  Montreuil-Bellay,  Maine-et-Loire. 

Dimensions 

D 

Wb 

Wh 

Wb:  Wh 

U 

MNHP  W ‘9’ 

112-3(100) 

36-4  (32) 

41-8(37) 

0-87 

- (-) 

MNHP'B’ 

111-7(100) 

- (-) 

41-0(37) 

— 

39  (35) 

SP,  de  Grossouvre 

160-0(100) 

51-0(32) 

60-0  (37) 

0-85 

53-0(33) 

Collection 

at  135-0(100) 

54-5  (40) 

58-0  (43) 

0-94 

44-5(33) 

SP,  Bourre 

120-0(100) 

40  (33) 

46-0  (38) 

0-87 

38-8(32) 

Description.  The  inner  whorls  of  this  species  are 

best  displayed 

by  the  specimen  from  Bourre  in  the 

Collections  illustrated  as  Plate  70,  figs.  3-5.  Up  to  a diameter  of  55  mm  the  coiling  is  relatively  evolute,  with  a 
medium-sized  umbilicus  (30%  of  diameter),  quite  shallow,  showing  on  the  mould  a rounded  and  undercut  wall. 


text-fig.  6.  Collignoniceras papale  (d’Orbigny).  Copies  of  d’Orbigny ’s  original  figures  ( 1 84 1 , pi.  1 09,  figs.  1 -3)  of 
the  holotype  from  the  ‘Craie  Chloritee  ou  Craie  Tuffeau  du  departement  de  Vaucluse’.  The  specimen  is  said  to  be 

120  mm  in  diameter. 


580 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  7.  Collignoniceras  papale  (d’Orbigny).  Adult  specimen  in  the  Sorbonne  Collections  (de  Grossouvre 
Collection),  from  either  Ponce  or  Bourre.  Reduced  x 0-6. 


The  intercostal  whorl  section  is  slightly  compressed  (Wb:Wh  is  0-9),  with  convergent  flanks,  broadly  rounded 
ventrolateral  shoulders  and  a flattened  venter.  The  costal  section  is  polygonal,  with  the  greatest  breadth  at  the 
umbilical  bulla.  There  are  thirteen  umbilical  bullae  per  whorl.  At  the  smallest  diameter  visible,  they  are  very 
elongate  and  lie  close  to  the  shoulder.  With  growth  the  maximum  development  migrates  outwards  leaving  a 
weak  development  only  at  the  umbilicus,  with  the  main  bulla  low  on  the  flank.  Broad,  strong,  straight,  slightly 
prorsiradiate  ribs  arise  from  the  bullae,  cross  the  flanks  and  connect  to  strong,  conical  inner  ventrolateral 
tubercles,  from  which  a broad,  strong  rib  sweeps  forwards  to  strong  outer  ventrolateral  clavi.  These  are  in  turn 
connected  to  elongate  siphonal  clavi  by  a broad,  low,  forwardly  directed  rib.  Between  long  ribs  there  are  some 
four  intercalatories,  usually  with  outer  ventrolateral  and  siphonal  clavi  only. 

From  50  mm  onwards  the  ribs  connecting  the  inner  and  outer  ventrolateral  tubercles  strengthen  and  at  55  mm 
they  have  fused  into  blunt,  oblique  clavi. 

During  middle  growth,  ornament  consists  of  distant,  weak  to  strong  umbilical  bullae,  displaced  progressively 
outwards  to  a low  or  even  mid  flank  position  (not  shown  on  d’Orbigny’s  figure),  which  give  rise  to  one  or  rarely  a 


EXPLANATION  OF  PLATE  69 

Figs.  1-2.  Collignoniceras  papale  ( d’Orbigny).  SP,  from  Bourre  (Loir-et-Cher)  (Saemann  Collection). 

Figs.  3-4.  Collignoniceras  woollgari  (Mantell);  BMNH  5742a-b,  paralectotypes  from  the  Middle  Chalk  near 
Lewes,  Sussex. 


PLATE  69 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


582 


PALAEONTOLOGY,  VOLUME  23 


pair  of  narrow,  straight,  prorsiradiate  ribs,  whilst  single  intercalated  ribs  arise  at  varying  levels  on  the  flank.  All 
ribs  bear  a pinched  ventrolateral  bulla  (if  weak)  or  horn  (if  strong).  These  are  commonly  limited  before  and 
behind  by  narrow  ribs,  which  loop  across  the  venter,  although  the  extent  of  this  looping  varies  widely  from 
specimens  in  which  it  predominates  (PI.  70,  fig.  4)  to  those  where  it  is  simple  (PI.  69,  fig.  1). 

Over  the  last  half  whorl  of  adult  body  chamber  the  tubercles  decline  markedly,  leaving  rather  weak,  relatively 
crowded  ribs  without  umbilical  bullae,  a weak,  oblique  to  radially  elongate  ventrolateral  tubercle  (which  may 
disappear  several  ribs  before  the  aperture)  and  a small  blunt  siphonal  tubercle  (text-fig.  7). 

The  suture  is  rather  simple,  with  a broad  E which  tapers  apically;  broad,  rather  simply  incised  and 
asymmetrically  bifid  E/L,  narrow  L and  smaller,  bifid  L/U2.  U2  is  small  (text-fig.  lc). 

Discussion.  The  material  before  us  shows  considerable  variation  in  the  relative  strength  of  umbilical 
bullae  and  ribs,  as  well  as  being  adult  (and  showing  typical  decline  in  ornament)  over  a range  of 
120-180  mm  diameter.  Nevertheless,  it  forms  a compact  species  group. 

Collignoniceras  canthus  is  immediately  distinguishable  on  the  basis  of  its  massively  tuberculate 
inner  whorls  and  feebly  ribbed,  almost  smooth  body  chamber  with  many  fine  ventral  clavi,  as 
discussed  on  p.  584.  There  are  closer  similarities  to  C.  turoniense,  but  here  the  massive  bullae  of  the 
inner  whorls  and  general  dominance  of  tuberculation  over  ribbing  is  diagnostic,  as  discussed  on 
p.  586. 

There  are  also  similarities  between  juveniles  of  C.  papale  and  C.  woollgari , but  papale  have  fewer, 
coarser  ribs  (compare  PI.  69,  figs.  3-4  and  PI.  70,  fig.  3),  with  strong  bullae  displaced  well  out  from  the 
umbilical  shoulder  and  much  more  prominent  inner  ventrolateral  tubercles.  C.  papale  in  middle 
growth  is  more  sharply  and  distantly  ribbed  and  does  not  have  the  prominent  ventrolateral  horn  of 
many  woollgari.  The  inner  and  outer  ventrolateral  tubercles  merge  into  pinched,  radially  elongated 
clavi  during  middle  growth  in  papale ; in  woollgari  they  are  distinct  to  a much  greater  size.  The  venter 
of  C.  papale  may  bear  strong  narrow  looped  ribs  at  a much  earlier  stage  than  woollgari  and  is  mature 
at  a much  smaller  diameter,  never  developing  the  spectacular  distantly  ribbed,  hypernodose  body 
chamber  of  the  latter. 

C.  carolinum  has  some  common  features,  particularly  its  rather  small  adult  size.  It  differs  in  having 
densely  and  evenly  ribbed  inner  whorls  without  strong  bullae,  and  a compressed  flat-sided  body 
chamber  without  the  umbilical  bullae,  strong  ventral  tubercles  and  broad  venter  with  looped  ribbing 
of  papale. 

C.  papaliforme  (Laube  and  Bruder)  (1887,  p.  237;  pi.  27,  figs.  3-4),  from  the  Turonian  Greensand 
of  the  White  Mountain,  near  Prague,  is  no  more  than  a deformed  C.  papale. 

Occurrence.  This  is  a relatively  long-ranging  species  in  the  Tuffeau  Blanc  of  Touraine,  first  appearing  in  the  St. 
Cyr-en-Bourg  Fossil  Bed  of  the  Saumur  region,  and  also  occurring  at  Montrichard,  Bourre,  and  Tourtenay 
(Deux  Sevres).  Elsewhere  in  France  there  are  records  from  Uchaux  (Vaucluse).  The  species  also  occurs  in  the 
Turonian  of  Czechoslovakia. 


Collignoniceras  canthus  (Sornay) 

Plate  73,  figs.  1-4 

1951  Ammonites  canthus  d’Orbigny  in  lift.;  Sornay,  p.  629,  text-figs,  le,  2. 

1955  Ammonites  ( Selwynoceras ) canthus  d’Orbigny  ms;  Sornay,  fiche  8,  figs.  1-2. 

1977  Collignoniceras  ( Selwynoceras ) canthus  (Sornay  ex  d’Orbigny  ms);  Hancock,  Kennedy  and 
Wright,  p.  156. 


EXPLANATION  OF  PLATE  70 

Figs.  1-2.  Collignoniceras  boreale  (Warren).  Cast  of  the  holotype,  Alberta  Museum  Collections  no.  CT  468, 
from  the  basal  beds  of  the  Smoky  River  Shale,  Grimshaw,  near  Peace  River,  Alberta. 

Figs.  3-5.  Collignoniceras  papale  (d’Orbigny),  nucleus,  showing  coarse  juvenile  ornament;  SP,  from  Bourre 
(Loir-et-Cher). 


PLATE  70 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


584 


PALAEONTOLOGY,  VOLUME  23 


Holotype.  By  monotypy  the  original  of  Sornay’s  (1951),  text-figs,  le,  2,  from  the  Tuffeau  Blanc  de  Touraine  of 
Bourre  (Loir-et-Cher),  Museum  d’Histoire  Naturelle,  Paris,  no.  6793. 

Dimensions  D wb  wh  wh:  wh  v 

Holotype 

MNHP6793  126(100)  40-8  (32)  49-5(39)  0-82  48-6(39) 

Description.  The  holotype  and  only  known  specimen  consists  of  the  internal  mould  of  a body  chamber  126  mm  in 
diameter  and  an  external  mould  of  the  umbilicus  of  the  inner  whorls.  The  umbilical  mould  shows  that  the  species 
bore  seven  massive  conical  umbilical  bullae  at  the  smallest  diameter  visible  (PI.  73,  fig.  3)  and  a similar  number  on 
the  following  whorl,  supplemented  by  three  ribs  lacking  bullae  but  extending  to  the  umbilicus.  From  the  bullae 
arose  rather  strong  ribs,  usually  in  pairs,  with  occasional  shorter  intercalated  ribs.  The  external  mould  of  the 
dorsum  of  the  last  part  of  the  phragmocone  shows  each  of  these  ribs  to  have  borne  a conical  ventral  tubercle 
whence  arose  a pair  of  feeble  ribs,  connecting  to  feeble  siphonal  tubercles  in  the  same  looped  style  seen  in 
Collignoniceras  papale  (d’Orbigny). 

The  body  chamber  shows  coiling  to  have  been  moderately  evolute,  with  a small  umbilicus  comprising  39%  of 
the  diameter.  The  umbilical  wall  is  low  and  rounded,  the  flanks  flattened  and  convergent,  with  a low  fastigiate 
venter  which  tends  to  become  rounded  towards  the  aperture.  The  maximum  whorl  breadth  is  low  on  the  flanks, 
close  to  the  umbilical  shoulder. 

On  the  early  part  of  the  body  chamber  there  are  weak  umbilical  bullae,  which  give  rise  to  pairs  of  low,  broad, 
radial  ribs,  almost  insensible  save  to  touch,  as  are  occasional  shorter,  intercalated  ribs.  The  ribs  become  pro- 
gressively finer,  denser  and  more  subdued  towards  the  mature  aperture,  and  are  gently  flexed. 

All  ribs  bear  faint,  low,  rounded  ventrolateral  clavi  which  give  rise  to  pairs  of  low  ribs  which  loop  forwards 
and  across  the  venter  to  low  siphonal  clavi  linked  into  a semi-continuous  serrated  ridge. 

The  rather  poorly  preserved  sutures  of  the  holotype  are  approximated,  confirming  it  as  an  adult. 

Discussion.  The  strongly  ornamented  inner  whorls  of  C.  canthus  place  it  in  the  same  group  as 
C.  papale  and  C.  turoniense.  It  differs  from  both  of  these  in  the  marked  decline  and  virtual 
disappearance  of  ornament  on  the  outer  whorl.  We  have  seen  no  intermediate  forms.  C.  carolinum 
(d’Orbigny)  has  delicately  and  densely  ribbed,  rather  than  coarsely  bullate  inner  whorls.  The  body 
chambers  of  the  two  are  more  similar,  especially  in  the  marked  decline  in  ornament,  but  carolinum  is 
much  more  compressed  and  flat-sided,  the  ribs  are  stronger,  with  quite  thick  ventral  development, 
and  stronger  siphonal  clavi. 

Occurrence.  C.  canthus  is  known  only  from  the  Tuffeau  Blanc  de  Touraine  of  Bourre. 

Collignoniceras  turoniense  (Sornay) 

Plate  71,  figs.  4-5;  Plate  72,  figs.  1-3 
1951  Prionotropis  turoniense  Sornay,  p.  630;  pi.  21,  figs.  1-3. 

1977  Collignoniceras  ( Selwynoceras ) turoniense  (Sornay);  Hancock,  Kennedy  and  Wright,  p.  156. 
Holotype.  MNHP  unregistered,  Peron  Collection,  from  Bourre  (Loir-et-Cher),  by  monotypy. 

Other  specimens  studied.  MNHP  ‘A’,  from  Bourre,  and  two  unregistered  specimens  in  the  de  Grossouvre 
Collection  (Sorbonne,  Paris),  probably  from  Bourre. 


Dimensions 

D 

Wb 

Wh 

Wb:  Wh 

U 

Holotype 

120  (100) 

48  (40) 

48-3  (40) 

1-0 

~ (-) 

MNHP  ‘A’ 

107  (100) 

52  (49) 

44  (41) 

118 

34-5(32) 

Sorbonne,  1 

125  (100) 

43-5  (35) 

48-5  (39) 

0-9 

34-5(28) 

at  107-5  (100) 

52-5  (49) 

43-5  (43) 

1-2 

23-0(21) 

EXPLANATION  OF  PLATE  71 

Figs.  1-3.  Collignoniceras  woollgari  (Mantell)  FSR  C273,  from  Ponce,  Sarthe. 

Figs.  4-5.  Collignoniceras  turoniense  (Sornay),  the  holotype,  MNHP,  Peron  Collection,  from  Bourre  (Loir-et- 
Cher). 


PLATE  71 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


586 


PALAEONTOLOGY,  VOLUME  23 


Description.  All  known  specimens  are  adults,  with  two-thirds  of  the  last  whorl  being  body  chamber,  and  none 
show  the  early  whorls.  Coiling  is  involute  on  the  phragmocone,  becoming  relatively  evolute  at  maturity,  with  a 
deep  umbilicus.  On  the  phragmocone  the  whorl  section  is  depressed,  with  convergent  flanks  and  a fastigiate 
venter  intercostally.  The  costal  section  is  even  more  depressed,  the  greatest  breadth  being  at  the  umbilical  bullae, 
and  subcarinate.  There  are  five  massive  blunt  conical  umbilical  nodes  per  whorl.  These  give  rise  to  groups  of  two 
or  three  broad,  low  ribs,  with  additional  ribs  intercalated  low  on  the  flank  between  the  groups.  At  the  smallest 
diameters  visible- these  bear  blunt  conical  inner  ventrolateral  tubercles  and  small  clavate  outer  ventrolaterals, 
with  a broad  low  rib  connecting  them  to  stronger  siphonal  clavi  borne  on  a blunt  keel.  On  the  last  part  of  the 
body  chamber  the  intercalated  ribs  decline,  the  inner  and  outer  ventrolateral  tubercles  combine  into  a blunt 
transversely  elongate  tubercle,  which  gives  rise  to  pairs  of  ribs  which  loop  to  strong  siphonal  clavi,  which  become 
first  rounded,  then  transversely  elongate.  Some  short  ventral  ribs  with  a siphonal  tubercle  are  intercalated,  to 
give  a serrated  blunt  keel;  there  are  three  to  five  siphonal  nodes  to  each  pair  of  umbilicals. 

On  the  body  chamber  the  umbilical  nodes  decline  in  strength  and  disappear  towards  the  aperture;  intercalated 
ribs  are  lost  and  the  primary  ribs  weaken  and  become  irregular  and  closely  spaced.  There  are  irregularly  spaced, 
clavate  ventrolateral  nodes,  which  also  decline  towards  the  aperture,  with  many  more  ventral  ribs  and  siphonal 
tubercles  than  ventrolateral. 

The  body  chamber  uncoils  markedly  and  the  shell  becomes  much  more  evolute  as  a result.  Whorl 
height:  breadth  ratio  decreases,  so  that  the  aperture  appears  relatively  constricted. 

None  of  the  specimens  shows  the  suture  well  but  they  appear  to  have  comprised  broad,  plump,  rather  simple 
bifid  lobes  and  saddles. 

Discussion.  The  inner  whorls  of  Collignoniceras  turoniense  are  easily  distinguished  from  those  of 
C.  woollgari  and  C.  carolinum,  which  are  densely  and  evenly  ribbed  by  comparison,  lacking  massive 
bullae.  In  middle  growth,  C.  turoniense  has  a much  more  massive  whorl,  broad  and  low  rather  than 
narrow  ribs  and  stronger  ventrolateral  than  umbilical  nodes.  The  adults  are  quite  distinct  (compare 
PI.  62,  figs.  1-2  and  PI.  71,  figs.  4-5). 

C.  canthus  has  similar  inner  whorls,  but  becomes  virtually  smooth  in  middle  and  later  growth, 
lacking  massive  umbilical  bullae  and  strong  ventrolateral  tubercles. 

C.papale  juveniles  (PI.  70,  figs.  3-5)  have  many  more  (typically  9-1 1)  and  smaller  umbilical  bullae, 
narrow  and  widely  spaced  ribs  and  more  markedly  clavate  ventrolateral  and  siphonal  tubercles.  In 
middle  and  later  growth  the  differences  between  the  two  lie  in  the  predominance  of  tuberculation  in 
C.  turoniense  and  of  ribbing  in  C.  papale,  the  latter  having  the  bullae  displaced  outwards  to  a lower 
flank  position  and  strong,  narrow,  well-differentiated  ventral  ribs  looping  between  the  ventrolateral 
and  siphonal  tubercles  with  intercalatories. 

C.  carolinum  is  compressed,  parallel-sided  and  feebly  ribbed  without  strong  bullae  in  middle  and 
later  growth. 

Occurrence.  C.  turoniense  is  known  only  from  the  Tuffeau  Blanc  de  Touraine  of  Bourre. 

Collignoniceras  boreale  (Warren) 

Plate  70,  figs.  1-2 

1930  Prionotropis  borealis  Warren,  p.  25,  pi.  3,  figs.  1-4;  pi.  4,  fig.  1. 

1940  Selwynoceras  borealis  Warren;  Warren  and  Stelck,  p.  151. 

Types.  The  holotype  is  the  original  of  Warren  1930,  pi.  3,  fig.  1,  University  of  Alberta  Museum  Collections 
no.  CT  468.  Paratypes  are  CT  469-76,  all  from  the  basal  beds  of  the  Smoky  River  Shale,  Grimshaw,  near  Peace 
River,  Alberta. 


EXPLANATION  OF  PLATE  72 

Figs.  1-3.  Collignoniceras  turoniense  (Sornay)  SP,  de  Grossouvre  Collection,  probably  from  Bourre  (Loir- 
et-Cher). 


PLATE  72 


KENNEDY,  wright  and  Hancock,  Collignoniceratid  ammonites 


588 


PALAEONTOLOGY,  VOLUME  23 


Description.  The  holotype,  a cast  of  which  is  before  us,  is  a slightly  distorted  mould  retaining  traces  of  shell  and 
consists  of  half  a whorl  of  body  chamber  and  one  quarter  of  a whorl  of  phragmocone  with  the  following 
dimensions: 

D Wb  Wh  Wb.Wh  U 

costal  92-5(100)  40  (43)  33-5  (36)  1 19  35-2(38) 

intercostal  90-2(100)  29-5(33)  31  (34)  0-98  35-2(39) 

Coiling  is  moderately  evolute,  the  umbilicus  comprising  38%  of  the  diameter,  broad  and  rather  shallow.  The 
umbilical  wall  slopes  gently  outwards  and  the  whorl  section  is  a compressed  oval  (whorl  breadth  to  height  ratio  is 
0-98)  with  flattened  flanks.  The  phragmocone  bears  three  long,  straight,  prorsiradiate  distant  ribs.  These  arise 
from  small  umbilical  bullae  and  also  bear  conical  inner  and  clavate  outer  ventrolateral  tubercles;  there  is  a 
siphonal  row  of  distant  clavi  corresponding  in  position  to  the  ventrolateral  tubercles.  Two  shorter,  intercalated 
ribs  are  also  present,  bearing  outer  ventrolateral  and  siphonal  clavi  only.  This  same  style  of  ventral  ornament  is 
shown  on  the  penultimate  whorl,  preserved  in  the  dorsum  of  the  body  chamber,  and  in  two  of  the  paratypes 
(Warren  1930,  pi.  3,  figs.  2-3). 

On  the  body  chamber  the  umbilical  bullae  decline  and  the  ribs  become  high,  distant,  and  flared  into  a 
ventrolateral  horn  which  supports  the  outer  ventrolateral  clavus.  There  is  a poorly  defined  siphonal  ridge, 
accentuated  into  siphonal  clavi,  and  the  upper  ventrolateral  and  siphonal  clavi  are  linked  by  broad  transverse 
ribs  which  show  incipient  doubling  with  a riblet  developing  at  both  front  and  rear. 

The  suture  is  simple  and  little  incised,  with  broad  bifid  saddles. 

Discussion.  Small  size  and  even  ventral  tuberculation  are  the  features  by  which  Warren’s  species  is 
most  easily  distinguished  from  C.  woollgari ; other  differences  are  noted  on  p.  572.  There  are  no  other 
species  with  which  it  is  likely  to  be  confused.  Of  interest,  however,  is  the  striking  resemblance  of  the 
holotype  to  specimens  of  C.  woollgari  from  the  Black  Hills  area  of  the  U.S.  Western  Interior,  which 
also  show  a very  even  and  equal  number  of  upper  ventrolateral  and  siphonal  clavi,  never,  apparently 
developing  the  intercalated  ribs  and  tubercles  of  what  we  take  as  typical  woollgari.  These  specimens 
(so  far  as  we  have  seen)  are  much  larger  when  adult  and  have  horns  with  a triangular  outline  in  ventral 
view  rather  than  flares.  These  Interior  examples  are  obviously  close  relatives  of  the  Canadian  form, 
although  their  precise  relative  ages  are  not  known. 

Occurrence.  As  for  types. 


Genus  lecointriceras  gen.  nov. 

Type  species.  Ammonites  fleuriausianus  d’Orbigny,  1841,  p.  350. 

Diagnosis.  Medium-sized,  involute  during  early  and  middle  growth,  becoming  evolute  at  maturity.  Whorls 
trapezoidal  when  young,  with  sparse  conical  umbilical  tubercles  giving  rise  to  pairs  of  low  broad  ribs,  with 
occasional  intercalatories.  All  ribs  bear  outer  ventrolateral  and  siphonal  clavi  on  a fastigiate  venter,  but  the 
appearance  and  persistence  of  inner  ventrolateral  tubercles  is  variable.  In  middle  growth  the  venter  often 
broadens  and  flattens,  the  ventrolateral  tubercles  fuse  into  a blunt  horn  and  there  is  a low  continuous  undulant 
siphonal  ridge,  strengthened  between  horns.  The  last  part  of  the  adult  body  chamber  is  contracted,  tubular  and 
unornamented  except  for  growth  lines,  and  the  aperture  is  simple. 

The  suture  is  simple  with  broad,  asymmetrically  bifid  E/L,  narrower  L and  smaller  bifid  L/U2. 

Discussion.  The  whorl  section,  massive  umbilical  tubercles  and  sparse  low  ribs  of  early  middle 
growth,  the  blunt  horns  and  the  tubular  body  chamber  distinguish  Lecointriceras  from  all  other 
collignoniceratids  and  the  persistence  of  short  ribs  on  the  sides  from  contemporaneous  Collignoni- 
ceras.  Some  C.  woollgari  develop  a short,  smooth  terminal  portion  to  the  body  chamber  but  their 


EXPLANATION  OF  PLATE  73 

Figs.  1-4.  Collignoniceras  canthus  (Sornay).  The  holotype,  SP  6793,  from  Bourre  (Loir-et-Cher).  3 is  the 
external  mould  of  the  nucleus;  4 shows  the  decline  in  ornament  over  the  last  part  of  the  body  chamber. 


PLATE  73 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


590 


PALAEONTOLOGY,  VOLUME  23 


compressed,  finely  ribbed  inner  and  middle  growth  stages,  much  narrower  flank  ribs,  retention  of 
multiple  siphonal  ribs  and  clavi  is  distinctive.  This  ventral  ribbing  and  retention  of  clavi  also 
distinguish  C.  canthus  and  C.  papale\  C.  turoniense  has  a smooth  body  chamber,  but  lacks  the 
massive  umbilical  tubercles  and  ventral  horns  in  middle  growth  and  on  the  first  part  of  the  body 
chamber.  The  phragmocone  of  some  Lecointriceras  and  the  adult  shell  of  C.  boreale  are  superficially 
similar,  but  Warren’s  species  has  compressed  finely  ribbed  inner  whorls  and  on  the  outer  whorl, 
which  is  slender  and  rounded  intercostally,  the  ribs  lack  a massive  bulla,  are  narrower  and  produced 
into  a narrow  flared  bituberculate  horn  rather  than  the  single  broad  protuberance  seen  in 
Lecointriceras. 

As  is  discussed  below.  Ammonites  vielbancii  d’Orbigny,  1850  is  a synonym  of  A.  fleuriausianus. 
Schliiter  (1871,  pp.  21-22)  believed  the  former  might  be  a synonym  of  Mammites  nodosoides 
(Schliiter),  and  Collignon  (1939)  and  Wiedmann  (1960,  1964)  referred  it  to  Mammites.  As 
Pervinquiere  (1907,  p.  31 1)  noted,  the  siphonal  tubercles  are  quite  distinctive. 

In  Europe  Lecointriceras  first  appears  in  the  mid-Turonian  St.  Cyr-en-Bourg  Fossil  Bed, 
accompanying  typical  Collignoniceras.  Its  origins  may  lie  in  one  of  the  undescribed  Thomelites-Mke 
forms  occurring  in  the  earliest  English  Turonian. 

Occurrence.  Widespread  in  the  French  Turonian  (Touraine  and  Aquitaine);  also  occurring  in  northern  Spain, 
Czechoslovakia,  north  Germany  and  southern  England. 


Lecointriceras  fleuriausianum  (d’Orbigny) 
Plate  74,  figs.  1-10;  Plate  75,  figs.  1-5;  text-figs.  8,  9 


1841  Ammonites  Fleuriausianus  d’Orbigny,  p.  350,  pi.  107,  figs.  1-3. 

1841  Ammonites  Woollgari  d’Orbigny,  p.  352  (pars),  pi.  108,  figs.  1-3. 

1850  Ammonites  Vielbancii  d’Orbigny,  p.  189. 

1860  Ammonites  Fleuriausianus  (d’Orbigny);  Courtiller,  p.  250,  pi.  3,  fig.  1. 

1867  Ammonites  Fleuriausianus  d’Orbigny;  Courtiller,  p.  7,  pi.  7,  figs.  1-4. 
non  1869  Ammonites  Fleuriauanus  d’Orbigny;  Schloenbach,  p.  291. 

1871  Ammonites  Vielbancii  d’Orbigny;  Schliiter,  p.  19  et  seq. 

?1872  Ammonites  Fleuriausianus  d’Orbigny;  Schliiter,  p.  28,  pi.  10,  figs.  1-3. 

1887  Acanthoceras  Fleuriausianum  d’Orbigny;  Laube  and  Bruder,  p.  234. 
non  1902  Acanthoceras  Fleuriausianum  d’Orbigny;  Petrascheck,  p.  147,  pi.  11,  figs,  la-b,  2. 

1907  Ammonites  Vielbancii  d’Orbigny;  Pervinquiere,  p.  311. 

1939  Mammites  Vielbancii  d’Orbigny;  Collignon,  p.  81,  pi.  11,  figs.  1,  2. 

1946  Ammonites  vielbancii  d’Orbigny;  Sornay,  p.  213. 

1946  Ammonites  fleuriausianus  d’Orbigny;  Sornay,  p.  214. 

1960  Mammites  vielbanci  (d’Orbigny);  Wiedmann,  p.  721 . 

1977  Collignoniceras  ( Selwynoceras ) fleuriausianum  (d’Orbigny);  Hancock,  Kennedy  and  Wright, 
p.  156. 


Type  series.  Ammonites  fleuriausianus  has  been  a poorly  understood  species,  although  the  type  figure  (if  taken  to 
be  natural  size)  is  an  accurate  representation  of  the  middle  growth  stages  and  the  type  series  survives.  In  his 
original  description  d’Orbigny  recorded  the  species  ‘en  place  dans  la  craie  chloritee  ou  craie  tufau  des  Martrous, 


EXPLANATION  OF  PLATE  74 

Figs.  1-10.  Lecointriceras  fleuriausianum  (d’Orbigny).  1 -2,  the  lectotype  of  ‘ Mammites ’ vielbancii  (d’Orbigny), 
MNHP  6779,  (d’Orbigny  Collection)  from  Saumur  (Maine-et-Loire).  3-5,  CS  629b,  from  the  environs  of 
Saumur  (Maine-et-Loire),  a juvenile  of  moderate  inflation.  6-7,  the  lectotype,  MNHP  6777b  (d’Orbigny 
Collection)  from  the  Calcaire  a Cephalopodes  of  Rochefort  (Charente-Maritime).  8-10,  FSM  125,  from 
Ponce,  Sarthe,  a hypernodose  juvenile. 


PLATE  74 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


*2$ 


592 


PALAEONTOLOGY,  VOLUME  23 


pres  de  Rochefort  (Charente-Inferieur);  M.  Dufrenoy  l’a  aussi  du  meme  lieu;  M.  d'Archiac  l’a  observee  a 
Gourdon  (Lot);  MM.  Dufrenoy  et  Graves  font  trouvee,  aux  environs  de  Saumur’  (d’Orbigny  1841,  p.  352).  In 
the  posthumous  catalogue  of  his  collection  (dating  from  1858-60)  the  following  are  recorded: 

6777  Saumur,  Maine-et-Loire,  3 specimens  (4  are  present). 

6777a  Martrous,  1 specimen  (missing). 

6777b  Rochefort,  Charente-Inferieur,  2 specimens  (3  are  present). 

6777c  Chatellerault,  Vienne,  2 specimens  (1  missing). 

The  Saumur  specimens  belong  to  at  least  two  species.  The  first,  34-5  mm  in  diameter,  is  a crushed  tuffeau 
specimen,  and  is  labelled  [La]  Fleche.  It  has  rather  flattened  flanks,  with  umbilical  bullae  giving  rise  to  2-3 
flexuous  ribs  with  some  intercalatories,  giving  a total  of  sixteen  ribs  per  whorl.  There  are  distinct  conical  inner 
ventrolateral  tubercles  and  subequal  outer  ventrolateral  and  siphonal  clavi,  which  show  it  to  be  afleuriausianum, 
as  is  a second  individual  with  an  estimated  diameter  of  55  mm,  but  having  little  indication  of  inner  ventrolateral 
tubercles  and  weak  siphonal  clavi. 

A third  specimen,  71  mm  in  diameter,  and  labelled  Saumur,  is  a worn,  wholly  septate  Jeanrogericeras 
reveliereanus.  The  final  specimen  has  ‘Rochefort’  written  on  it  in  pencil  and  is  also  a J.  reveliereanus,  with  a 
diameter  of  104  mm.  Superficially  it  could  be  the  basis  of  d’Orbigny’s  side  view  but  it  lacks  all  signs  of  a 
siphonal  clavus. 

Two  specimens  from  Rochefort  are  associated  with  a plaque  labelled  6777b.  Both  are  well  preserved  on  one 
side,  the  larger  55  mm  in  diameter,  the  smaller  35  mm,  and  appear  to  be  part  of  d’Orbigny’s  original  suite.  The 
larger  of  these,  the  most  typical  in  the  series,  is  here  designated  lectotype. 

The  single  specimen  to  survive  of  those  originally  labelled  6777c  is  a very  battered,  crushed,  distorted  specimen 
in  yellow  tuffeau.  Umbilical  bullae  give  rise  to  pairs  of  ribs,  terminating  in  rounded  ventral  clavi,  with  no  sign  of 
siphonal  nodes,  suggesting  it  to  be  a mammitid  or  Jeanrogericeras.  Chatellerault  was  not  mentioned  as  a locality 
by  d’Orbigny  in  his  original  description  and  thus  this  specimen  is  not  a syntype. 

The  types  of  A.  vielbancii,  herein  regarded  as  a synonym,  also  present  a confused  situation.  It  is  a Prodrome 
species  introduced  (d’Orbigny  1850,  p.  189,  no.  11)  as  follows:  ‘ Vielbancii , d’Orb.,  Paleont.,  1,  p.  352,  pi.  108, 
figs.  1-3.  Sous  le  faux  nom  de  Woolgarii,  Mantell.  Martrous,  Saumur,  Tourtenay.’ 

In  Paleontologie  Franqaise  (1841,  p.  354)  he  cites  the  species  as  occurring  more  widely,  but  we  take  these 
references  (which  include  England)  to  be  to  the  true  Collignoniceras  woollgari. 

The  d’Orbigny  catalogue  lists  the  following: 

6779  Saumur,  Maine-et-Loire,  3 (4  specimens). 

6779a  Bords  de  la  Vienne,  2 (1  missing). 

6770b  Rochefort,  (illegible)  (missing). 

whilst  d’Orbigny  notes  that  his  lateral  view  (pi.  108,  fig.  1)  is  of  a specimen  in  his  collection  and  the  apertural 
view  is  of  a specimen  in  the  Ecole  des  Mines. 

Inspection  shows  that  the  d’Orbigny  specimens  have  become  mixed.  The  Rochefort  specimen  is  present,  but 
labelled  6779.  It  is  poorly  preserved,  but  may  be  the  basis  of  d’Orbigny’s  side  view.  The  specimen  from  the  Bords 
de  la  Vienne  is  not  a syntype;  it  is  a large  Mammites  nodosoides.  As  Sornay  has  discussed  (1946,  p.  214),  the 
specimen  figured  in  side  view  by  d’Orbigny  does  not  look  like  any  of  the  poor  specimens  which  survive  in  his 
collections  under  the  name  vielbancii,  and  certainly  there  is  little  resemblance  between  d’Orbigny’s  figures  and 
the  specimen  re-figured  by  Collignon  as  ‘type’— which  we  take  to  be  a valid  lectotype  designation.  Even  the 
specimen  in  the  School  of  Mines  upon  which  d’Orbigny  (1841,  p.  354)  said  his  apertural  view  is  based  (no.  A35.3, 
locality  unknown:  ‘Bassin  de  la  Loire,  achete  de  Stur’  reads  the  label)  does  not  correspond  to  the  figure  (compare 
text-figs.  8 A-c  and  9 a-b).  We  would  suggest,  in  fact,  that  the  illustrations  are  composite,  the  side  view  being 
based  on  the  poor  Rochefort  specimen  of  appropriate  size,  combined  with  the  ornament  of  the  huge  Mammites 
no.  6779a  from  the  ‘Bords  de  la  Vienne’,  the  apertural  view  being  based  on  the  School  of  Mines  specimen  plus  the 
Mammites. 

Description.  The  smallest  individuals  we  have  seen  are  approximately  30  mm  in  diameter.  At  this  size  the  coiling 
is  fairly  involute  (umbilicus  = 25%  or  less  of  diameter)  and  the  umbilicus  quite  deep,  with  a rounded  wall.  The 


EXPLANATION  OF  PLATE  75 

Figs.  1-5.  Lecointriceras  fleuriausianum  (d’Orbigny).  1-3,  FSM  120,  4-5,  FSM  121,  compressed  and  inflated 
middle-aged  individuals  from  the  Turonian  of  Sarthe. 


PLATE  75 


Kennedy,  wright  and  HANCOCK,  Collignoniceratid  ammonites 


text-fig.  8.  A-c,  copies  of  d’Orbigny’s  original  figures  of  ‘ Ammonites  Woollgari  Man  tell’  (1841,  pi.  108,  figs.  1-3) 
= Ammonites  vielbancii  d’Orbigny,  1850.  The  illustration  is  said  in  the  text  to  be  reduced  by  a third  and  on  the 
plate  by  a half,  d-f,  copies  of  d’Orbigny’s  original  figures  of  Ammonites  fleuriausianus  (1841,  pi.  107,  figs.  1-3). 
The  illustration  is  said  to  be  reduced  by  a third. 


text-fig.  9.  Lecointriceras fleuriausianum  (d’Orbigny)  a,  B.  EMP  A35.3,  ‘Bassin  de  la  Loire,  achete  de  Stur’ — the 
original  of  d’Orbigny’s  (1841)  pi.  108,  fig.  2.  Reduced  x 0-66.  c,  d.  FSM  1 19,  an  adult  from  Ponce,  Sarthef?) 
showing  the  smooth,  tubular  termination  to  the  body  chamber.  Reduced  x 0-6  approx. 


596 


PALAEONTOLOGY,  VOLUME  23 


intercostal  whorl  section  is  typically  compressed,  with  the  greatest  breadth  low  on  the  convergent  flank  and  with 
rounded  shoulders  and  venter.  In  the  costal  section  the  greatest  breadth  is  at  the  umbilical  bulla  and  whorl 
breadth  to  height  ratios  vary  greatly  up  to  1:2,  with  concave  inner  flanks  and  a fastigiate  venter. 

Ornament  consists  of  weak  to  strong  conical  umbilical  bullae,  7-9  per  whorl,  giving  rise  to  pairs  of  low,  broad 
straight  ribs,  with  occasional  intercalated  ribs  arising  low  on  the  flank.  The  ribs  decline  somewhat  in  strength  on 
the  mid-flank  but  then  strengthen  into  rounded  inner  ventrolateral  tubercles.  These  are  connected  by  a 
strengthened  rib  to  strong  clavate  outer  ventral  tubercles,  from  which  a broad  subdued  rib  sweeps  forwards  to 
a subequal  clavate  siphonal  tubercle. 

This  general  style  of  ornament  varies  from  individual  to  individual,  with  slender,  feebly  bullate  forms  with 
weak  ribs  (PI.  74,  figs.  6-7)  and  strongly  bullate  inflated  forms  with  strong  ribs  (PI.  74,  figs.  8-10).  In  many 
individuals,  including  the  lectotype,  there  are  no  inner  ventrolateral  tubercles  below  diameters  of  35-42  mm; 
occasionally  they  do  not  appear  until  55  mm. 

From  50  mm  onwards  there  is  usually  a change  in  ornament;  the  bullate  umbilical  tubercles  become  more 
distant,  the  associated  ribs  lower  and  broader,  effaced  at  mid-flank  in  some  specimens.  There  are  usually  7-9 
bullae  and  16-22  ribs  per  whorl.  The  outer  ventral  tubercles  weaken  rapidly  and  disappear;  at  the  same  stage  the 
inner  ventrolateral  tubercles  strengthen  without  joining  the  weakening  ventral  tubercles  (PI.  77,  fig.  4).  The 
former  inner  ventrolateral  tubercles  gradually  develop  into  strong  to  massive  horns  on  the  shoulder,  triangular 
when  viewed  ventrally  and  relatively  narrow  when  viewed  laterally,  developed  both  upwards  and  outwards. 
At  this,  the  ‘ vielbancii'  stage,  the  venter  becomes  relatively  broad,  with  a continuous  low  undulant  siphonal 
ridge,  strengthened  between  horns  at  what  corresponds  to  the  site  of  the  now  coalesced  siphonal  clavi.  The 
shell  now  closely  resembles  a Mammites  in  all  but  the  siphonal  ridge. 

This  style  of  ornament  extends  onto  the  first  half  of  the  adult  body  chamber,  by  which  stage  the  siphonal  ridge 
may  become  very  reduced  (text-fig.  9 c-d).  On  the  last  half  of  the  body-chamber,  extending  for  just  over  a 
quarter  whorl,  all  ribs  and  tubercles  are  lost  and  there  is  a relatively  smooth,  compressed  and  constricted 
terminal  portion  with  convergent  sides,  broadly  rounded  shoulders  and  a flattened  venter,  ornamented  only  by 
low,  prorsiradiate  growth  striae.  The  aperture  is  simple  and  entire. 

The  suture  line  is  relatively  simple,  with  a broad  medial  element  to  E;  broad,  asymmetrically  bifid  E/L; 
narrow,  symmetrically  bifid  L;  smaller  asymmetrically  bifid  L/U2;  and  small  and  narrow  U2. 

Discussion.  D’Orbigny ’s  original  figure  is  idealized  and  bears  little  relationship  to  the  surviving 
syntypes  in  his  collection;  in  his  explanation  of  the  plate  he  says  the  figure  is  reduced  by  a third,  so 
that  the  specimen  is  far  larger  than  the  proposed  lectotype,  being,  presumably,  the  Martrous 
specimen  which  is  now  lost.  The  lectotype  agrees  well  with  the  dimensions  given  by  d’Orbigny  for  his 
smaller  specimen  (1841,  p.  350).  Juveniles  of  this  species  vary  in  the  strength  of  the  umbilical 
tubercles;  the  lectotype  is  worn  but  was  probably  a slender,  weakly  tuberculate  variant.  This 
variation  continues  into  middle  growth,  where  both  slender  and  robust  individuals  are  known  (PI.  74, 
figs.  3-10). 

The  striking  contracted  tubular  termination  of  the  body-chamber  of  adults  of  this  species  occurs  at 
disparate  sizes.  Most  specimens  we  have  seen  appear  to  be  juveniles  of  individuals  that  would  have 
been  adult  at  approaching  1 50  mm  diameter,  but  a specimen  in  the  collections  at  Le  Mans  is  complete 
at  only  100  mm,  with  half  a whorl  of  the  body  chamber  so  modified.  Unfortunately  our  sample  of 
adults  is  too  small  to  show  if  the  species  shows  a size  dimorphism. 

Some  of  the  early  references  to  this  species  are  doubtful.  Schloenbach’s  (1869)  material  probably 
belongs  to  Barroisiceras,  whilst  Schluter’s  specimen  (1872,  p.  28;  pi.  10,  figs.  1-3),  if  indeed  a true 
L.  fleuriausianum,  has  suffered  great  post-mortem  crushing  to  give  a very  compressed  whorl  section. 

Lecointriceras  carinatum  sp.  nov.,  described  below,  differs  from  L.  fleuriausianum  in  its  smaller 
adult  size,  early  loss  of  umbilical  tubercles  and  ribs,  together  with  retention  of  a fastigiate  venter  on 
the  adult  body  chamber,  which  bears  an  undulose  siphonal  and  flanking,  semi-continuous  lateral 


EXPLANATION  OF  PLATE  76 

Figs.  1-2.  Collignoniceras  carolinum  (d’Orbigny),  MNE1P  W8,  an  adult  body-chamber  from  an  unknown 
locality  in  the  Tuffeau  Blanc  de  Touraine. 

Figs.  3-5.  Lecointriceras  carinatum  sp.  nov.  The  holotype,  EMP,  Ponce(?),  Sarthe. 


PLATE  76 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


598 


PALAEONTOLOGY,  VOLUME  23 


keels  formed  by  coalescence  of  ventral  and  siphonal  clavi.  Differences  from  L.  costatum  sp.  nov.  are 
discussed  below. 

The  combined  features  of  L.fleuriausianum  as  here  described  are  so  distinctive  that  confusion  with 
any  other  collignoniceratid  is  unlikely.  Juveniles  have  a passing  similarity  to  some  Barroisiceratinae; 
species  of  Barroisiceras  have  less  prominent  umbilical  tubercles  and  many  strong,  narrow  ribs  at  a 
comparable  size;  whilst  Forresteria  and  similar  genera  have  an  additional,  lateral  row  of  tubercles.  In 
middle  growth  there  is  a superficial  resemblance  to  Mammites,  but  that  genus  never  develops  a 
siphonal  tubercle. 

Occurrence.  This  species  is  common  at  the  level  of  the  mid-Turonian  St.  Cyr-en-Bourg  Fossil  Bed  in  the  Saumur 
area  in  Touraine,  occurs  in  northern  Aquitaine,  Vaucluse,  Provence,  northern  Spain,  north  Germany(?),  and 
Devon,  England. 


Lecointriceras  carinatum  sp.  nov. 

Plate  76,  figs.  3-5 

Holotype.  A body-chamber  in  the  Collections  of  the  School  of  Mines,  Paris,  labelled  Ponce(?)  and  in  pencil 
Choffaticeras ’ typique;  ’’Thomasites"' . It  is  clearly  from  either  Ponce  or  Bourre. 

Description.  The  holotype  and  only  known  specimen  is  a half  whorl,  largely  body-chamber  and  in  typical  rather 
coarse  tuffeau  preservation.  Coiling  is  very  involute  with  a tiny  umbilicus  (10%  of  diameter).  The  dorsum  of  the 
specimen  (PI.  76,  figs.  3-5)  shows  the  whorl  section  of  the  inner  whorls  to  have  been  slightly  depressed,  with  the 
greatest  breadth  at  the  umbilical  shoulder,  concave,  convergent  flanks  and  a fastigiate  venter.  There  were  sparse 
umbilical  bullae  giving  rise  to  low,  broad  ribs  which  terminate  at  elongate  ventrolateral  clavi,  with  a sharp 
siphonal  keel,  accentuated  into  clavi  which  correspond  to  the  ventrolaterals. 

On  the  first  part  of  the  body  chamber  ornament  is  similar.  There  are  low  broad  flank  ribs  which  terminate  in 
long  clavi  linked  into  undulant  keels,  flanking  a similarly  undulant  keel  in  which  clavi  merge  towards  the 
aperture. 

The  poorly  preserved  suture  shows  a typical  broad  bifid  E/L,  narrow  L,  and  broad  L/U2,  all  with  only  minor 
incisions. 

Discussion.  The  single  known  individual  is  so  distinctive  that  erection  of  a new  species  is  justified.  The 
inner  whorls  are  typical  of  a Lecointriceras,  differing  from  L.fleuriausianum  in  the  sparse,  low,  broad 
ribs  and  presence  of  keels.  Absence  of  a quadrate-whorled  vielbancii  stage  makes  the  body  chamber 
equally  distinctive.  There  is  a striking  similarity  to  Masiaposites  Collignon,  1965,  a late  Turonian 
form  best  known  from  Madagascar  and  currently  regarded  as  a vascoceratid;  however  its  siphonal 
keel  is  entire  and  its  sutures  are  much  more  deeply  incised,  rather  like  that  of  Neoptychites,  and  the 
siphonal  keel  continuous  throughout  ontogeny. 

Occurrence.  The  species  is  known  only  from  the  type  occurrence  at  Ponce(?),  Sarthe  (mid-Turonian). 

Lecointriceras  costatum  sp.  nov. 

Plate  77,  figs.  1-3 

1902  Acanthoceras  Fleuriausianum  d’Orbigny;  Petrascheck,  p.  147,  pi.  11,  figs.  1-2. 

Holotype.  AM  55  from  the  Tuffeau  Blanc  of  Saumoussay,  Maine-et-Loire,  France. 

Other  specimens  studied.  AM  22  from  Montsoreau,  Maine-et-Loire;  AM  53,  54,  60,  101,  and  102  from 
Saumoussay,  Maine-et-Loire,  France. 


EXPLANATION  OF  PLATE  77 

Figs.  1-3.  Lecointriceras  costatum  sp.  nov.  1-2,  the  holotype,  AM  55,  from  Saumoussay,  Maine-et-Loire; 
3,  AM  53  from  Saumoussay,  Maine-et-Loire. 

Fig.  4.  Lecointriceras  fleuriausianum  (d’Orbigny).  AM  36  from  Saumoussay,  Maine-et-Loire;  oblique  view  to 
show  the  concurrent  weakening  of  the  outer  and  the  strengthening  of  the  inner  ventrolateral  clavi. 


PLATE  77 


Kennedy,  wright  and  Hancock,  Collignoniceratid  ammonites 


600 


PALAEONTOLOGY,  VOLUME  23 


D 

Wb 

Wh 

Wb:  Wh 

U 

R 

AM  55  (Holotype) 

125-5 

(100) 

~ (— ) 

53  + 

33 

c.  21 

AM  53 

95 

(100) 

c.  36  ( ) 

c.  44-5 

0-81 

AM  60 

183 

(100) 

71 

54 

0-76 

AM  101 

165 

(100) 

14 

at  129 

(100) 

50-5 

56 

0-90 

35 

17 

AM  102 

109 

(100) 

40 

45 

0.89 

30 

Description.  This  is  a moderately  evolute  and  relatively  compressed  Lecointriceras,  with  the  greatest  whorl- 
breadth  still  at  the  umbilical  tubercles  in  costal  section.  Of  the  fourteen  to  twenty-one  ribs  slightly  less  than  half 
are  long;  the  shorter  ribs  start  about  halfway  up  the  sides.  Each  long  rib  bears  an  umbilical  bulla,  a clavus  high  on 
the  sides  rather  than  in  the  normal  position  of  an  inner  ventrolateral,  an  outer  ventrolateral  clavus  and  a siphonal 
clavus.  The  siphonal  clavi  are  elevated  above  the  shoulder  clavi  and  up  to  a diameter  of  125  mm  may  form  a 
nodose  keel.  During  the  earlier  ontogeny  the  high  lateral  clavi  are  weaker  than  those  on  the  shoulders,  but  at 
diameters  which  may  be  anything  from  70-110  mm  the  upper  lateral  clavi  strengthen  and  the  shoulder  clavi 
weaken;  the  upper  lateral  clavi  eventually  become  ventrolateral  horns  on  the  body-chamber.  Similarly  the 
umbilical  bullae  become  weak  and  are  not  present  on  all  long  ribs  beyond  diameters  of  100  mm.  The  adult  body- 
chamber  begins  at  about  125  mm  diameter,  but  none  of  the  specimens  seen  has  well-preserved  sutures. 

Discussion.  L.  costatum  differs  from  L.  fleuriausianum  in  having  a more  compressed  whorl  section 
with  flatter  sides,  weaker  umbilical  tubercles  (which  are,  however,  still  stronger  than  in  typical 
Collignoniceras  spp.),  siphonal  clavi  elevated  above  the  outer  ventrolateral  clavi  and  persistent  outer 
ventrolateral  and  upper  lateral  clavi  through  much  of  ontogeny,  certainly  from  a diameter  of  40  mm 
to  about  125  mm. 

Occurrence.  All  known  French  specimens  are  from  the  mid-Turonian  Tuffeau  Blanc  of  the  Saumur  region.  In 
that  formation  ammonites  are  most  common  in  the  St.  Cyr-en-Bourg  Fossil  Bed,  but  we  have  not  found  any 
specimens  of  L.  costatum  ourselves;  as  Amedro  and  Badillet  (1978)  have  pointed  out,  ammonites  do  occur  at 
other  levels  in  the  Tuffeau  Blanc.  The  specimens  figured  by  Petrascheck  were  from  Labiatus-Planer  at  Leubnitz 
and  Briessnitz  near  Dresden  in  the  German  Democratic  Republic. 


EVOLUTIONARY  AND  STRATIGRAPHIC  CONCLUSIONS 

The  origins  of  Collignoniceras  and  the  Collignoniceratidae  seem  to  lie  in  late  Thomelites  of 
Acanthoceratidae,  the  transition  involving  a raising  of  the  mid-venter  and  forwards  displacement  of 
siphonal  clavi  and  ribs  to  give  a ventral  chevron  ornament.  This  is  indicated  by  a few  scraps  we  have 
seen  from  the  Cenomanian-Turonian  boundary  beds  in  Devon.  Lecointriceras  may  also  arise  in  this 
way,  or  be  a slightly  later  offshoot  from  already  distinct  Collignoniceras : the  low  Turonian  record  is 
too  poor  to  be  certain.  In  the  United  States  C.  woollgari  overlaps  late  Mammites  nodosoides  (W.  A. 
Cobban,  in  litt.);  in  Europe  C.  woollgari  and  L.  fleuriausianum  co-occur  in  the  earliest  of  the  French 
Tuffeau  faunas.  C.  woollgari  is  a long-ranging  species  which  occurs  throughout  the  mid-Turonian 
zone  of  which  it  is  the  index  species.  In  Europe  we  have  detected  no  evolutionary  changes  in  the 
successive  Collignoniceras  faunas  studied.  In  contrast,  W.  A.  Cobban’s  work  on  western  interior 
sequences  allows  recognition  of  an  early  form,  in  which  both  long  and  short  ribs  persist  in  middle  and 
later  growth,  and  a late  form  in  which  long  ribs  dominate.  That  this  is  not  seen  in  Europe  suggests  that 
typical  individuals  had  reached  the  U.S.  Western  Interior  by  the  beginning  of  woollgari  Zone  time, 
and  underwent  subsequent  local  differentiation  which  did  not  occur  in  European  populations.  The 
other  collignoniceratids  described  here  are  mostly  long  ranging:  C.  carolinum,  C.  papale  and 
L.  fleuriausianum  range  through  most  of  the  woollgari  Zone.  L.  costatum  is  restricted  to  the  lower 
part,  L.  carinatum,  C.  turoniense  and  C.  canthus  to  middle  and  low  upper  levels  in  the  Zone. 

These  disappointingly  meagre  stratigraphic  conclusions  mean  that  any  subdivision  of  the  broad 
woollgari  Zone  must  be  based  on  other  groups.  We  have  already  suggested  that  a local  sequence  of 
Romaniceras  can  be  used  in  Touraine:  R.  (R.)  kallesi  (oldest)  ->  R.  ( Yubariceras ) ornatissimum  -*• 
R.  (R.)  deverianum  (youngest)  (Hancock,  Kennedy  and  Wright  1977;  Kennedy,  Wright  and 


KENNEDY  ET  AL .:  COLLIGNONICERATID  AMMONITES 


60 : 


Hancock,  this  volume).  The  lower  two  of  these  are  clearly  correlated  with  the  xvoollgari  Zone,  but  we 
are  not  entirely  certain  whether  R.  deverianum  marks  a level  at  the  very  top  of  the  woollgari  Zone  or  at 
the  base  of  the  succeeding  Subprionocyclus  neptuni  Zone.  Ammonites  are  too  scarce  at  this  level  in 
both  England  and  northern  France  for  us  to  be  sure  either  way;  Romaniceras  appears  to  be  absent 
from  the  rich  neptuni  Zone  fauna  of  the  Chalk  Rock  (Wright  1979)  but  occurs  in  the  Uchaux 
(Vaucluse)  faunas. 

Acknowledgements.  We  are  grateful  to  the  following  colleagues  for  allowing  us  to  examine  specimens  in  their 
care,  and/or  for  much  useful  discussion:  Dr.  J.  Sornay,  Dr.  D.  Pajaud,  the  late  General  M.  Collignon,  Dr.  R. 
Busnardo,  Dr.  J.  P.  Lefranc,  Dr.  J.  Lovail,  Mr.  M.  Gruet,  Professor  K.  Young,  Dr.  C.  Duerdon,  Dr.  M.  R. 
Cooper,  Dr.  W.  A.  Cobban,  Dr.  W.  A.  Popenoe,  Dr.  E.  G.  Kauffman,  Dr.  V.  Housa,  Dr.  R.  Zazvorka, 
Dr.  M.  K.  Howarth,  Mr.  D.  Phillips,  Professor  T.  Matsumoto  and  Dr.  I.  Hayami.  The  financial  support  of  the 
Royal  Society,  British  Association  for  the  Advancement  of  Science  and  N.E.R.C.  is  gratefully  acknowledged  by 
Kennedy  and  Hancock.  We  thank  the  staff  of  the  Geological  Collections,  University  Museum,  Oxford,  and 
of  the  Department  of  Geology,  King’s  College,  London  for  their  help  and  assistance. 


REFERENCES 

adkins,  w.  s.  1931.  Some  Upper  Cretaceous  ammonites  in  western  Texas.  Univ.  Tex.  Bull.  3101, 35-72,  pis.  2-5. 
amedro,  f.  and  badillet,  G.  1978.  Repartition  des  ammonites  dans  quelques  coupes  du  Turonien  des  environs  de 
Saumur  (Maine-et-Loire).  C.r.  hebd.  Seanc.  Acad.  Sci.,  Paris,  (D)  286,  323-325. 
arkhanguelsky,  A.  D.  1912.  The  fossil  fauna  from  the  shores  of  the  Aral  Sea.  1:  Upper  Cretaceous  deposits, 
part  1.  Izv.  Turkest.  Otd.  imp.  russ.  geogr.  Obshch.  8 (2).  [In  Russian.] 

— 1916.  The  molluscs  of  the  Upper  Cretaceous  of  Turkestan.  Trudy  geol.  Komit.  S.-Peterburg,  152, 1-57  + vi, 
pis.  1-8.  [In  Russian.] 

arkell,  w.  J.  et  al.  1957.  Cephalopoda  Ammonoidea.  Treatise  on  Invertebrate  Paleontology , L,  Mollusca  4, 
490  + xxii  pp. 

Basse,  e.  1948.  Quelques  ammonites  nouvelles  du  Cretace  de  Colombie  (Am.  Sud).  Bull.  Soc.  geol.  Fr.  (5)  18, 
691-698,  pi.  25. 

— 1950.  Quelques  mollusques  du  Cretace  de  Colombie.  Bull.  Soc.  geol.  Fr.  (5)  20,  245-255,  pi.  11. 

bose,  e.  1928  (mis-dated  1927).  Cretaceous  ammonites  from  Texas  and  northern  Mexico.  Univ.  Tex.  Bull.  2748, 
143-357,  pis.  1-18. 

breistroffer,  m.  1947.  Notes  de  nomenclature  paleozoologiques.  Proc.  verb.  mens.  Soc.  Sci.  Dauphine , 26  (195), 
5 pp. 

cobban,  w.  A.  and  scott,  G.  R.  1972.  Stratigraphy  and  ammonite  fauna  of  the  Graneros  Shale  and  Greenhorn 
Limestone  near  Pueblo,  Colorado.  Prof.  Pap.  U.S.  geol.  Surv.  645,  108  + v pp.,  41  pis. 
collignon,  M.  1939.  Fossiles  cenomaniens  et  turoniens  du  Menabe  (Madagascar).  Annls.  geol.  Serv.  Mines 
Madagascar,  10,  59-105,  pis.  1-11. 

— 1965.  Atlas  des  fossiles  caracteristiques  de  Madagascar  (Ammonites),  12  (Turonien),  82+iv  pp., 
pis.  376-413.  Tananarive,  Service  Geologique. 

courtiller  1860.  Description  de  trois  nouvelles  especes  d’ammonites  du  terrain  cretace.  Memoires  de  la 
Societe  imperiale  d' Agriculture  Sciences  et  Arts  d' Angers.  3rd  series,  3,  246-252,  pis.  1-3. 

— 1867.  Les  Ammonites  du  tuffeau.  Annales  de  la  Societe  Linneenne  du  Departement  de  Maine-et-Loire,  9th 
year,  pp.  1-8,  pis.  1-8. 

diener,  c.  1925.  Ammonoidea  neocretacea.  Fossilium  Catalogus  1:  Animalia,  29,  244  pp. 

fieber,  f.  x.  1853.  Synopsis  der  europaischen  Orthopteren  mit  besonderes  Riicksicht  der  Bohmischen  Arten. 

Lotos,  3,  90-104,  115-129,  138-154,  168-176,  184-188,  201-207,  232-238,  252-261. 
fritsch,  a.  1872.  Cephalopoden  der  bohmischen  Kreideformation.  52  pp.,  16  pis.  Prag,  Fr.  Rivnac. 
geinitz,  H.  b.  1871-1875.  Das  Elbthalgebirge  in  Sachsen.  Part  1,  Der  untere  Quader.  Palaeontographica,  20, 
319  pp.,  67  pis. 

gray,  d.  A.  1965.  The  strati  graphical  significance  of  electrical  resistivity  marker  bands  in  the  Cretaceous  strata  of 
the  Leatherhead  (Fetcham  Mill)  borehole,  Surrey.  Bull.  geol.  Surv.  Gr.  Br.  23,  65-1 15,  pis.  8-9. 
grossouvre,  a.  de.  1894  (mis-dated  1893).  Recherches  sur  la  craie  superieur.  2:  Paleontologie— Les  ammonites 
de  la  craie  superieur.  Mem.  Serv.  Carte  geol.  det.  Fr.  264  + (ii)  pp.,  39  pis. 
haas,  o.  1946.  Intraspecific  variation  in,  and  ontogeny  of,  Prionotropis  woollgari  and  Prionocyclus  wyomingensis. 
Bull.  Am.  Mus.  not.  Hist.  86,  141-224,  pis.  11-24. 


602 


PALAEONTOLOGY,  VOLUME  23 


Hancock,  j.  m.,  Kennedy,  w.  j.  and  wright,  c.  w.  1977.  Towards  a correlation  of  the  Turonian  sequences  of 
Japan  with  those  of  north-west  Europe.  Spec.  Pap.  palaeont.  Soc.  Japan , 21,  151-168. 
hattin,  D.  e.  1975.  Stratigraphy  and  depositional  environment  of  Greenhorn  Limestone  (Upper  Cretaceous)  of 
Kansas.  Bull.  Kans.  Univ.  geol.  Surv.  209,  128  + (iii)  pp.,  10  pis. 
hoepen,  E.  c.  n.  van.  1955.  Turonian-Coniacian  ammonites  from  Zululand.  S.  Afr.  J.  Sci.  51,  361-377. 
kullmann,  j.  and  wiedmann,  J.  1970.  Significance  of  sutures  in  phylogeny  of  Ammonoidea.  Paleont.  Contr. 
Univ.  Kans.  47,  32  pp. 

laube,  G.  c.  and  bruder,  g.  1887.  Ammoniten  der  bohmischen  Kreide.  Palaeontographica,  33,  217-239, 
pis.  23-29. 

logan,  w.  N.  1898.  The  invertebrates  of  the  Benton,  Niobrara  and  Fort  Pierre  Groups.  Univ.  geol.  Surv.  Kans.  4 
(Paleont.)  (Upper  Cretaceous  8),  431-518,  pis.  86-120. 

mantell,  G.  1822.  The  fossils  of  the  South  Downs;  or  illustrations  of  the  geology  of  Sussex.  327  + xvii  pp.,  42  pis. 
London,  Lupton  Relfe. 

matsumoto,  T.  1959.  Upper  Cretaceous  ammonites  of  California,  part  II.  Mem.  Fac.  Sci.  Kyushu  Univ. 
(D.  Geol.),  Special  Vol.  1,  1-172,  pis.  1-41. 

— 1965.  A monograph  of  the  Collignoniceratidae  from  Hokkaido,  Part  1.  Mem.  Fac.  Sci.  Kyushu  Univ. 
(D.  Geol.),  16,  1-80,  pis.  1-18. 

— 1971.  A monograph  of  the  Collignoniceratidae  from  Hokkaido,  Part  5.  Ibid.  (D.  Geol.),  21,  129-162, 
pis.  21-24. 

— and  wright,  c.  w.  1966.  Collignoniceras  Breistroffer,  1947  (Mollusca,  Ammonoidea):  application  to  place 
on  the  official  list  of  generic  names  in  zoology  with  priority  from  1876.  Bull.  zool.  Nom.  23,  57-59. 

meek,  F.  B.  1876.  A report  on  the  invertebrate  Cretaceous  and  Tertiary  fossils  of  the  upper  Missouri  country.  In 
hayden,  F.  v.  Report  of  the  United  States  Geological  Survey  of  the  Territories,  9,  629  + lxiv  pp.,  45  pis. 
orbigny,  a.  d\  1840-1842.  Paleontologie  Franqaise.  Description  des  Mollusques  rayonnes  fossiles.  Terrains 
Cretaces,  1 (Cephalopodes),  662  pp.,  151  pis.  Paris,  Victor  Masson. 

— 1850.  Prodrome  de  Paleontologie  stratigraphique  universelle  des  animaux  mollusques  et  rayonnes  . . .,  2, 
428  pp.  Paris,  Victor  Masson. 

pervinquiere,  L.  1907.  Etudes  de  paleontologie  tunisienne  I.  Cephalopodes  des  terrains  secondaires.  Carte  geol. 
Tunisie,  438  + v pp.,  27  pis. 

petrascheck,  w.  1902.  Die  Ammoniten  der  sachsischen  Kreideformation.  Beitr.  Palaont.  Geol.  Ost.-Ung.  14, 
131-162,  pis.  7-12. 

pictet,  f.  J.  and  campiche,  G.  1858-1860.  Description  des  fossiles  du  terrain  cretace  des  environs  de  Sainte- 
Croix,  part  2 (1),  Description  des  fossiles.  Materiaux  pour  la  paleontologie  suisse  (2)  2,  29-380,  pis.  1-43. 
Geneva,  J.  Kessmann  and  H.  Georg  for  F.  J.  Pictet. 

powell,  j.  d.  1963.  Turonian  (Cretaceous)  ammonites  from  northeastern  Chihuahua,  Mexico.  J.  Paleont.  37, 
1217-1232,  pis.  166-171. 

RAWSON,  P.  F.,  CURRY,  D.,  DILLEY,  F.  C.,  HANCOCK,  J.  M.,  KENNEDY,  W.  J.,  NEALE,  J.  W.,  WOOD,  C.  J.  and 

worssam,  B.  c.  1978.  A correlation  of  Cretaceous  rocks  in  the  British  Isles.  Spec.  Rep.  geol.  Soc.  Lond.  9, 70  pp. 
schloenbach,  u.  1869.  Bemerkungen  fiber  einige  Cephalopoden  der  Gosaubildungen.  Jb.  K.-K.  geol. 
Reichsanst.,  Wien,  19,291-294. 

schluter,  c.  1871-1876.  Cephalopoden  der  oberen  deutschen  Kreide.  Palaeontographica,  21,  1-120,  pis.  1-35; 
24,3-143  -t-xii,  pis.  36-55. 

sharpe,  d.  1853-1857.  Description  of  the  fossil  remains  of  Mollusca  found  in  the  Chalk  of  England.  Cephalo- 
poda. Palaeontogr.  Soc.  ( Monogr .),  68  pp.,  27  pis. 

shimizu,  s.  1 932.  On  a new  type  of  Senonian  ammonite,  Pseudobarroisiceras  nagaoi  Shimizu  gen.  et  sp.  nov.  from 
Teshio  Province,  Hokkaido.  Jap.  J.  Geol.  Geogr.  10,  1-4,  pi.  1. 
sornay,  j.  1946.  Remarques  sur  deux  especes  de  d’Orbigny:  Ammonites  vielbancii  et  A.  fleuriausianus.  Bull. 
Mus.  Hist.  nat.  Paris.  (2)  18,  213-216. 

— 1951.  Sur  deux  especes  d’ammonites  inedites  de  d’Orbigny  et  sur  une  espece  nouvelle  du  tuffeau  de 
Touraine.  Bull.  geol.  Soc.  Fr.  (6)  1,  627-631,  pi.  21. 

— 1955.  Ammonites  ( Selwynoceras ) canthus  d’Orbigny.  Palaeont.  univers.  (n.s.)  8. 
sowerby,  j.  de  c.  1828.  The  mineral  conchology  of  Great  Britain  . . .,  6,  pis.  581-597. 

spath,  l.  f.  1923.  On  the  ammonite  horizons  of  the  Gault  and  contiguous  deposits.  Summ.  Progr.  geol.  Surv. 
Lond.  for  1922,  139-149. 

— 1926.  On  new  ammonites  from  the  English  Chalk.  Geol.  Mag.  63,  77-83. 

warren,  p.  s.  1930.  Three  new  ammonites  from  the  Cretaceous  of  Alberta.  Trans.  R.  Soc.  Can.  (3)  24  (4),  21-26, 
pis.  1-4. 


KENNEDY  ET  AL COLLIGNONICERATID  AMMONITES 


603 


warren,  p.  s.  and  stelck,  c.  r.  1940.  Cenomanian  and  Turonian  faunas  in  the  Pouce  Coupe  District,  Alberta 
and  British  Columbia.  Trans.  R.  Soc.  Can.  (3)  34  (4),  143-152,  pis.  1-4. 

Wedekind,  r.  1916.  Uber  Lobus,  Suturallobus  und  Inzision.  Zentbl.  Miner.  Geol.  Palaont.  (B)  1916  (8), 
185-195. 

wiedmann,  j.  1960  (mis-dated  1959).  Le  Cretace  superieur  de  l’Espagne  et  du  Portugal  et  ses  cephalopodes. 

C.r.  Congr.  Soc.  sav.-Dijon  1959:  Colloque  sur  le  Cretace  superieur  franqais,  709-764. 
windmoller,  t.  1882.  Die  entwickelung  des  Plaeners  im  nordwestlichen  theile  des  Teutoburger  Waldes  bei 
Lengerich.  Jb.  preuss.  geol.  Landesanst,  BergAkad.  [2]  (for  1881),  Abh.  3-54,  pi.  19. 
wright,  c.  w.  1979.  The  ammonites  of  the  English  Chalk  Rock.  Bull.  Br.  Mus.  nat.  Hist.  (Geol.),  31,  281-332, 
7 pis. 

— and  Kennedy,  w.  J.  In  juignet,  p.,  Kennedy,  w.  J.  and  wright,  c.  w.  1973.  La  limite  Cenomanien-Turonien 
dans  la  region  du  Mans  (Sarthe):  stratigraphie  et  paleontologie.  Annls.  Paleont.  (Invert.),  59,  207-242. 

— and  matsumoto,  T.  1954.  Some  doubtful  Cretaceous  ammonite  genera  from  Japan  and  Saghalien.  Mem. 
Fac.  Sci.  Kyushu  Univ.  (D.  Geol.),  4,  107-134. 

— and  wright,  e.  v.  1951.  A survey  of  the  fossil  Cephalopoda  of  the  Chalk  of  Great  Britain.  Palaeontogr.  Soc. 
(Monogr.),  40  pp. 

young,  K.  1963.  Upper  Cretaceous  ammonites  from  the  Gulf  Coast  of  the  United  States.  Univ.  Tex.  Pubis.  6304, 
373  + viii  pp.  (82  pis.). 


W.  J.  KENNEDY 
C.  W.  WRIGHT 
University  Museum 
Parks  Road,  Oxford  OX1  3PW 
and  Wolfson  College,  Oxford  OX2  6UP 

Typescript  received  28  March  1979 
Revised  typescript  received  8 November  1979 


J.  M.  HANCOCK 

Department  of  Geology 
King’s  College,  Strand 
London  WC2R  2LS 


THE  TRILOBITE  ECCOPTOCHILE  FROM  THE 
ORDOVICIAN  OF  NORTHERN  PORTUGAL 


Abstract.  The  eccoptochilinid  trilobite  fauna  from  the  Ordovician  of  the  Valongo  area,  north  Portugal,  is 
revised.  The  holotype  of  Eccoptochile  (1  Eccoptochile)  mariana  (Verneuil  and  Barrande,  1855)  is  redescribed  and 
figured  and  the  species  is  restricted  to  the  type  specimen  and  two  specimens  from  Valongo.  Specimens  pre- 
viously described  as  E.  (IE.)  mariana  from  Spain,  north  Portugal,  and  southern  England,  together  with  other 
and  new  material  from  Portugal  are  here  included  within  the  new  species  E.  ( Eccoptochile ) almadenensis. 
E.  ( Eccoptochile ) cf.  clavigera  (Beyrich,  1845)  is  recorded  from  the  Valongo  area. 

This  revision  of  the  genus  Eccoptochile  from  the  Ordovician  of  north  Portugal  forms  part  of  a 
larger  project  concerned  with  systematic  description  and  distribution  studies  of  the  Ordovician 
trilobite  faunas  of  that  region.  The  faunas  from  the  Valongo  area  about  10  km  east  of  Porto  (text- 
fig.  1)  have  been  well  known  since  Delgado  published  extensive  faunal  lists  from  the  beds  (1908, 
pp.  106-109);  only  ‘ Uralichas  Ribeirof  (Delgado,  1892,  1897)  was  described.  Delgado  listed 
‘ Cheirurus  claviger  Beyrich’,  ‘ Cheirurus  Guillieri  Tromelin  (aff.  C.  clanger  Beyrich)’,  and  ''Cheirurus 
sp.  n.  (aff.  C.  Sedgwicki  McCoy)’  from  his  uppermost  division,  the  ‘Schistes  a Uralichas  Ribeirof , 
of  the  ''Ordovician  moyen'  from  the  Valongo  area.  Prior  to  this  Sharpe  (1849)  had  recorded 
‘ Chirurus ’ from  the  Porto  region  but  this  specimen  was  later  recognized  by  Salter  (1853)  as 
‘ Placoparia  Zippei,  Boeck’.  The  most  recent  systematic  work  on  this  group  was  by  Curtis  (1961) 
who  apparently  regarded  all  three  of  the  species  listed  by  Delgado  as  conspecific  and  referred  them 
to  Eccoptochile  mariana  (Verneuile  and  Barrande). 

The  ‘Schistes  a Uralichas  Ribeirof  have  generally  been  regarded  as  Llandeilo  in  age  (Costa  1931; 


by  M.  ROMANO 


text-fig.  1.  Simplified  geological  map  of  the  area 
south  of  Valongo  (after  Delgado  1908),  showing 
localities  (asterisks)  where  the  species  of  Eccoptochile 
described  in  the  text  have  been  recorded. 


IPalaeontology,  Vol.  23,  Part  3,  1980,  pp.  605-616,  pis.  78-79.) 


606 


PALAEONTOLOGY,  VOLUME  23 


Teixeira  1955;  Thadeu  1956)  and  more  recent  work  on  certain  elements  of  the  fauna,  notably 
harpids  (Romano  1975),  placopariids  (Romano  1976),  and  dionidids  (Henry  and  Romano  1978), 
suggests  a possible  Lower  Llandeilo  age,  equivalent  to  the  Placoparia  ( Coplacoparia ) tournemini 
biozone  of  Spain  and  Brittany  (Hammann  1971a;  Henry  and  Clarkson  1975).  The  ‘Schistes  a 
Uralichas  Ribeirof  are  included  within  the  upper  part  of  the  Valongo  Formation  (Romano  and 
Diggens  1973-1974)  which  is  a thick  sequence  of  argillaceous  sediments  ranging  in  age  from  Upper 
Llanvirn  ( Didymograptus  murchisoni  Zone)  to  ?Upper  Llandeilo.  The  formation  crops  out  about 
10  km  east  of  Porto  and  it  is  from  this  area  that  the  bulk  of  the  collections  were  made  by  Delgado, 
Wattison,  and  the  present  author  with  J.  N.  Diggens.  The  problem  of  accurately  locating  the  material 
collected  by  Wattison  was  outlined  earlier  (Romano  1976)  and  similar  difficulties  arise  with  some  of 
the  specimens  from  the  Delgado  collection. 

The  collections  used  in  this  paper  are  housed  in  the  British  Museum  (Natural  History),  London 
(Wattison  Collection — (BM  In));  Ecole  Nationale  Superieure  des  Mines,  Paris  (T);  Servigos 
Geologicos,  Lisbon  (Delgado  Collection— SG);  Institute  of  Geological  Sciences  (GSM),  and  in  the 
Geology  Department,  University  of  Sheffield  (SU). 


SYSTEMATIC  PALAEONTOLOGY 

General  remarks.  E.  ( Eccoptochile ) clavigera  (Beyrich,  1 845),  E.  (I  Eccoptochile)  mariana  (Verneuil 
and  Barrande,  1855)  and  E.  (? Eccoptochile)  guillieri  (Tromelin  in  Guillier,  1873)  form  a relatively 
homogeneous  group  within  which  the  north  Portuguese  specimens  clearly  belong.  The  first  two  are 
generally  regarded  as  valid  species  but  E.  (IE.)  guillieri  has  fairly  recently  been  placed  into  synonymy 
with  E.  (IE.)  mariana  by  Hammann  (1974,  p.  105).  E.  (IE.)  guillieri  was  compared  with 
E.  (E.)  clavigera  by  Tromelin  and  Lebesconte  (1876,  p.  637)  who  noted  that  the  glabella  of  the 
former  differed  from  that  in  E.  (E.)  clavigera  in  being  smooth,  more  convex,  with  the  posterior  end 
of  the  axial  furrows  curved  inwards  more  strongly.  The  outline  diagrams  and  locality  information 
of  the  specimens  of  E.  (IE.)  guillieri  shown  in  text-fig.  2 (g,  h,  i)  were  kindly  sent  to  me  by  Dr.  J.-L. 
Henry;  they  are  of  the  holotype  (2g)  and  a topotype  (2h,  i from  the  Kerforne  collection).  The  latter 
is  an  incomplete  but  undeformed  specimen,  preserved  in  a nodule,  from  the  type  locality  ‘la  Butte 
du  Creux’,  near  Saint-Denis-d’Orques  (Sarthe);  Dr.  Henry  informed  me  that  it  is  Llanvirn  or 
Llandeilo  in  age.  This  topotype  shows  a very  narrow  (sag.)  frontal  area  and  a strongly  and  evenly 
curved  glabella  in  lateral  view.  From  these  two  specimens  E.  (IE.)  guillieri  warrants  retention  as  a 
separate  species  and  is  treated  as  such  in  this  paper. 

The  most  commonly  reported  species  of  Eccoptochile  in  Iberia  and  the  Armorican  Massif  is 
E.  mariana  (Curtis,  op.  cit.;  Hammann  1971,  1974;  Lindstrom,  Racheboeuf  and  Henry  1974),  but  a 
recent  study  of  the  holotype  of  this  species  by  the  author  suggests  that  the  species  has  been  interpreted 
too  widely  in  the  past.  The  holotype  of  mariana  is  redescribed  and  figured  here. 

Prantl  and  Pribyl  (1948)  erected  the  subgenus  Eccoptochile  ( Eccoptochiloides ) on  the  basis  of  the 
thorax  containing  only  ten  segments  and  the  four  pairs  of  pleural  spines  on  the  pygidium.  As  the 
thorax  and  pygidium  of  E.  (IE.)  mariana  are  unknown  the  subgeneric  status  of  mariana  is  still  in 
doubt. 

The  morphological  terms  used  are  essentially  those  listed  by  Harrington  et  al.  (in  Moore,  1959). 
Lateral  glabellar  lobes  and  furrows  are  labelled  ‘L’  and  ‘S’  respectively  and  are  numbered  from  the 
posterior  forwards.  The  classification  employed  is  that  of  Henningsmoen  (in  Moore,  1959)  and  Lane 
(1971). 


Family  cheiruridae  Hawle  and  Corda,  1847 
Subfamily  eccoptochilinae  Lane,  1971 
Genus  eccoptochile  Hawle  and  Corda,  1847 


Type  species.  Cheirurus  claviger  Beyrich,  1845 


ROMANO:  ORDOVICIAN  TRILOBITE  ECCOPTOCHILE 


607 


text-fig.  2.  Outline  sketches  of  the  cephala  or  cranidia  of  the  holotype  and  other  material  of 
selected  species  of  Eccoptochile : a-c,  Eccoptochile  (Eccoptochile)  almadenensis  sp.  nov.  a,  b, 
Holotype  (selected),  from  Hammann,  1974,  pi.  12,  fig.  192c  and  192b  (reversed  for  comparison); 
c,  from  Curtis,  1961,  pi.  2,  fig.  1.  d-f , Eccoptochile  ( Eccoptochile ) clavigera  (Beyrich);  d,  Holo- 
type, from  Beyrich,  1845,  pi.  (unnumbered),  fig.  2;  e,  /,  from  Barrande,  1852,  pi.  40,  figs.  1,  2. 
g-i,  Eccoptochile  (? Eccoptochile)  guillieri  Tromelin  in  Guillier);  g,  Holotype,  h,  i,  Topotype. 
Both  drawings  taken  from  photographs  and  drawings  supplied  by  Dr.  J.-L.  Henry. y-/,  Eccopto- 
chile (? Eccoptochile)  mariana  (Verneuil  and  Barrande);  y,  k , Holotype,  from  Verneuil  and 
Barrande,  1855,  pi.  23,  fig.  4 and  present  paper,  pi.  1,  figs.  1-4;  /,  Paratype,  from  Curtis, 
1961,  pi.  1,  fig.  1 and  refigured  here,  pi.  1,  figs.  5,  6.  Sketches  drawn  to  approximately  the  same 

size. 


Eccoptochile  (? Eccoptochile)  mariana  (Verneuil  and  Barrande,  1855) 

Plate  78,  figs.  1-7;  text-fig.  2 j-1 

*1855  Cheirurus  marianus  Verneuil  and  Barrande,  p.  970,  pi.  23,  fig.  4 (not  p.  972,  pi.  28  as  stated  by 
Hammann,  1974,  p.  105). 

1961  Eccoptochile  mariana  (Verneuil  and  Barrande);  Curtis,  p.  6,  pi.  1,  fig.  1 (not  pi.  1,  fig.  2,  pi.  2; 
figs.  1,  2,  ?pl.  3,  fig.  1). 

1974  Eccoptochile  cf.  mariana  (Verneuil  and  Barrande);  Hammann,  p.  105  (referring  to  Curtis,  1961, 
pi.  1,  fig.  1). 

Diagnosis.  (Modified  from  Verneuil  and  Barrande,  1855,  p.  970.)  A species  of  Eccoptochile  with  the  following 
characteristics:  strongly  arched  glabella  with  evenly  curved  longitudinal  profile  and,  with  occipital  ring  vertical, 
highest  part  level  with  the  anterior  part  of  L2.  Wide  frontal  area  over  12%  of  the  glabellar  length  (excluding 
occipital  ring)  and  consists  of  a more  or  less  flat  preglabellar  field  and  a gently  rounded  anterior  border. 
Palpebral  lobe  level  with  the  posterior  part  of  L2  to  the  posterior  part  of  L3.  Eye  ridges  are  faintly  visible 
running  from  the  anterior  of  the  palpebral  lobe  towards  S3.  Hypostoma,  thorax,  and  pygidium  unknown. 

Type  and  figured  material.  Holotype:  T 150  (Plate  78,  figs.  1-4).  Internal  mould  of  incomplete  cranidium 
(Verneuil  and  Barrande,  1855,  pi.  23,  fig.  4).  Other  figured  material.  BM  In49177  (Plate  78,  figs.  5,  6)  (Curtis, 
1961,  pi.  1,  fig.  1);  BM  In49182  (Plate  78,  fig.  7). 


608 


PALAEONTOLOGY,  VOLUME  23 


Horizon  and  locality.  Holotype  from  ‘Puente  de  las  Ovejas’  near  Ciudad  Real,  Spain;  Upper  Llandeilo 
(Hamman,  1974,  p.  105).  BM  In49177  and  In49182  from  Covelo,  near  Valongo,  north  Portugal;  upper  part 
of  Valongo  Formation,  probably  Lower  Llandeilo. 

Description  of  holotype.  Measurements  with  occipital  ring  vertical:  length  (sag.)  of  glabella  (excluding  occipital 
ring)  and  frontal  area,  15-75  mm;  length  of  glabella,  14  00  mm.  Glabella  longer  than  wide  with  even,  out- 
wardly curved  lateral  margins,  slightly  indented  at  S3,  and  a broadly  rounded  anterior  margin;  widest  part 
of  the  glabella  just  anterior  to  the  S2  furrows.  LI  lobes  subtriangular  in  outline,  about  one-quarter  glabellar 
length  and  delimited  by  deep,  well-marked  SI  furrows  which  have  an  S-shaped  trace  and  die  out  just  under 
one-third  glabellar  width  from  axial  furrows.  L2  lobes  rectangular  in  outline,  shorter  (trans.)  than  LI  and 
about  the  same  length  (exsag.).  S2  furrows  shorter  and  less  well-marked  than  SI,  evenly  curved,  parallel  to  the 
abaxial  part  of  SI,  starting  just  posterior  to  the  midlength  of  the  glabella.  L3  similar  in  shape  and  orientation 
to  L2,  but  appear  to  be  very  slightly  longer.  S3  furrows  parallel  to  S2  but  do  not  reach  as  far  towards  the 
midline.  S3  start  at  nearly  two-thirds  the  glabellar  length  from  the  posterior  margin. 

Glabella  strongly  arched  transversely  with  a subtriangular  cross  section.  Longitudinally  (occipital  ring 
vertical)  the  glabella  is  evenly  curved  dorsally,  highest  part  lying  above  the  anterior  part  of  L2.  Median 
glabellar  lobe,  L2  and  L3  without  independent  convexity  but  LI  lobes  are  slightly  bulbous.  Frontal  area  wide 
(sag.,  exsag.),  of  more  or  less  constant  width  around  the  frontal  glabellar  lobe  but  increasing  at  anterolateral 
corners  where  anterior  margin  of  fixed  cheek  turns  back  rather  sharply  to  give  a more  angular,  although  still 
rounded  outline.  Frontal  area  consists  of  an  inner  preglabellar  field  which  is  more  or  less  flat  or  very  slightly 
upwardly  concave  which  grades  into  the  frontal  lobe  of  the  glabella  without  a marked  furrow.  Preglabellar 
field  also  grades  into  anterior  border  which  is  gently  rounded  and  lies  horizontally.  Anterolaterally  the  border 
is  slightly  wider.  At  anterolateral  comers  border  appears  to  be  directed  more  upwards  but  this  may  be  an 
effect  of  deformation.  Axial  furrows  well-marked  from  the  occipital  furrow  to  S3  where  there  is  a deep  pit 
just  abaxial  to  axial  furrow.  Anterior  to  this  pit  axial  furrow  rapidly  dies  out.  Occipital  furrow  curved  forwards 
behind  the  median  glabellar  lobe  and  where  it  runs  into  the  axial  furrows,  deep  posterior  to  the  LI  lobes  and 
wide  and  shallow  in  the  median  part.  Occipital  ring  not  complete:  posterior  to  the  LI  lobes  ring  curves 
forwards.  Incomplete  free  cheeks  are  narrow  (trans.)  opposite  the  palpebral  lobes  and  fairly  flat.  Posterior 
border  furrow  deep,  starting  from  the  axial  furrow  just  posterior  to  the  occipital  furrow.  Posterior  border 
narrow  (exsag.)  and  convex.  Convex  (tr.)  palpebral  lobe  slightly  curved,  lying  oblique  to  sagittal  line  and 
separated  from  fixed  cheek  by  a well-marked  palpebral  furrow  which  dies  out  anteriorly  along  length  of  lobe. 
Faint  eye  ridge  extends  from  palpebral  lobe  to  axial  furrow  at  S3.  Palpebral  lobe  level  with  posterior  part  of  L2 
to  the  posterior  part  of  L3.  Faint  granular  ornament  on  glabella  but  the  distribution  is  not  clear.  On  the 
fixed  cheeks  there  is  an  irregular  distribution  of  pits. 

The  figured  material  from  Valongo  assigned  to  this  species  is  virtually  identical  to  the  holotype,  differing 
mainly  in  convexity.  The  Portuguese  specimens  are  flattened  dorso-ventrally  and  slightly  distorted  obliquely. 
The  transverse  and  longitudinal  profiles  of  the  cranidia  do  not  show  the  high  convex  glabella  of  the  holotype 
but  the  relative  proportions  of  the  cranidia  are  the  same.  This  species  is  discussed  further  below. 


EXPLANATION  OF  PLATE  78 

Figs.  1-7.  Eccoptochile  (?.  Eccoptochile)  mariana  (Verneuil  and  Barrande).  1-4,  holotype,  internal  mould; 
T 150.  ‘Puente  de  las  Ovejas’  near  Ciudad  Real,  Spain;  Upper  Llandeilo.  1-3,  dorsal,  frontal,  lateral  views 
respectively,  x 3.  4,  detail  of  cheek  ornament,  x9.  5,  6,  internal  mould;  In49177.  Covelo,  north  Portugal. 
Upper  part  of  Valongo  Formation;  Lower  Llandeilo.  5,  dorsal  view.  6,  frontal  view.  Approximately  x 2. 
7,  internal  mould;  In49182.  Covelo,  north  Portugal.  Upper  part  of  Valongo  Formation;  Lower  Llandeilo. 
Dorsal  view,  x 1. 

Figs.  8,  9.  Eccoptochile  {Eccoptochile)  almadenensis  sp.  nov.  Internal  moulds.  8,  GSM  CR  1526.  Gorran 
Quartzites,  Perhaver  Beach,  Cornwall;  Llandeilo.  Dorsal  view,  x2.  9,  SG  3A2.  1400  m S 32°  E of  Covelo 
church,  north  Portugal.  Upper  part  of  Valongo  Formation;  Lower  Llandeilo.  Dorsal  view,  x 1. 


PLATE  78 


romano,  Ordovician  trilobite  Eccoptochile 


610 


PALAEONTOLOGY,  VOLUME  23 


Eccoptochile  ( Eccopotchile ) almadenensis  sp.  nov. 

Plate  78,  figs.  8,  9;  Plate  79,  figs.  1-7;  text-fig.  2 a-c 

1896  Cheirurus  ( Eccoptocheile ) marianus  (De  Verneuil);  Reed,  p.  164. 

1907  Cheirurus  sedgwicki  M'Coy;  Lake  in  Reid,  p.  39. 

1908  Cheirurus  claviger  Beyrich;  Delgado,  ? p.  57  (refigured  by  Thadeu,  1947,  pi.  3,  fig.  2),  ? p.  80, 

p.  106. 

1908  Cheirurus  guillieri  Trom.  (aff.  C.  claviger  Beyr.);  Delgado,  p.  106. 

1908  Cheirurus  sp.  n.  (aff.  C.  sedgwicki  McCoy);  Delgado,  p.  106. 

1916  Eccoptochile  mariana  (Verneuil  and  Barrande);  Barton,  p.  106. 

* 1918  Cheirurus  claviger  var.  marianus  Verneuil  and  Barrande  emend.  Born;  Born,  p.  351,  pi.  27,  fig.  1. 
1947  Cheirurus  claviger  Beyrich;  Thadeu,  p.  228,  pi.  3,  fig.  3. 

1958  Eccoptochile  clavigera  (Beyrich);  Whittard,  p.  115  (specimen  from  Perhaven  Beach,  Cornwall). 
1961  Eccoptochile  mariana  (Verneuil  and  Barrande);  Curtis,  p.  6,  pi.  1,  fig.  2 ( non  fig.  1),  pi.  2,  figs.  1,  2, 
pi.  3,?  fig.  1. 

1969  Eccoptochile  ( Eccoptochile ) sp.  indet;  Racheboeuf,  p.  74,  pi.  2,  figs.  3 a,  b. 

19716  Eccoptochile  marianus  (Verneuil  and  Barrande);  Hammann,  pp.  267,  270. 

1974  Eccoptochile  clavigera  (Beyrich)?;  Sadler,  p.  73. 

1974  Eccoptochile  (. Eccoptochile ) mariana  (Verneuil  and  Barrande);  Lindstrom,  Racheboeuf,  and 
Henry,  ? pp.  20,  21. 

1974  Eccoptochile  mariana  (Verneuil  and  Barrande);  Hammann,  p.  105,  text-fig.  39,  pi.  11,  figs.  188- 
191,  pi.  12,  figs.  192-198. 

1978  Eccoptochile  mariana  (Verneuil  and  Barrande);  Henry  and  Romano,  p.  335. 

Diagnosis.  (Modified  from  Hammann,  1974,  p.  106.)  Species  of  Eccoptochile  with  glabella  strongly  convex, 
anterior  lobe  descending  almost  vertically  to  preglabellar  field.  Frontal  area  relatively  narrow  (sag.);  anterior 
border  steeply  upturned  forming  an  angle  with  lateral  borders  of  free  cheeks  (viewed  dorsally).  Eyes  start 
approximately  level  with  S2  and  reach  back  to  SI.  Fixed  cheeks  narrow  (sag.).  Anterior  thoracic  segments 
pointed,  becoming  gradually  more  rounded  posteriorly.  Internal  surface  of  exoskeleton  smooth  except  for  pits 
on  cheeks. 

Type  and  figured  material.  Holotype:  (SMG  X 337a)  Internal  mould  of  cephalon  with  seven  thoracic  segments 
(figured  Born,  1918,  p.  351,  pi.  27,  fig.  1;  Hammann,  1974,  p.  105,  pi.  12,  figs.  192  a-c).  Paratypes:  (BM 
In49 178-80)  Curtis,  1961,  p.  6,  pi.  1,  fig.  2,  pi.  2,  figs.  1 and  2 respectively;  (SMF  24779-82,  24783^3,  24784, 
24785a,  24787)  Hammann,  1974,  p.  105,  pi.  11,  figs.  188,  ?189,  190-191,  pi.  12,  figs.  193-198.  Other  material: 
GSM  GR  1526;  SG  1704,  SG  171 1,  and  three  unnumbered  specimens  in  drawer  labelled  3A2  in  SG  (figured  here 
PI.  78,  fig.  9,  PI.  79,  figs.  6,  7). 

Horizons  and  locality.  Holotype  from  Valdemosillo,  approximately  16  km  ENE  of  Almaden,  Spain;  Upper 
Llandeilo.  Paratypes.  BM  In49178-80  from  Covelo,  near  Valongo,  north  Portugal;  upper  part  of  Valongo 
Formation,  probably  Lower  Llandeilo.  SMF  24779,  24785a  from  Corral  de  Calatrava  (near  Ciudad  Real, 
Spain);  Co  Illf,  Upper  Llandeilo:  SMF  24780-82,  from  Corral  de  Calatrava;  Co  Hie,  Upper 


EXPLANATION  OF  PLATE  79 

Figs.  1 -7.  Eccoptochile  ( Eccoptochile ) almadenensis  sp.  nov.  1 -6,  internal  moulds,  7,  external  impression.  Upper 
part  of  Valongo  Formation;  Lower  Llandeilo.  1-4,  SG  1704.  1650  m S 20°  W of  the  summit  of  Santa  Justa, 
Valongo,  north  Portugal.  5,  SG  1711,  6,  7 (both  in  drawer  labelled  3A2),  1400  m S 32°  E of  Covelo 
church,  north  Portugal.  1 -3,  dorsal,  lateral,  frontal  views  respectively,  x 2;  4,  detail  of  thoracic  segment,  x 4. 
5-7,  dorsal  views,  x 1,  x0-75,  x 1 respectively. 

Fig.  8.  Eccoptochile  (? Eccoptochile)  cf.  mariana  (Verneuil  and  Barrande).  Internal  mould;  (no  number,  same 
box  as  SG  1704).  1650  m S 20°  W of  the  summit  of  Santa  Justa,  Valongo,  north  Portugal.  Upper  part  of 
Valongo  Formation;  Lower  Llandeilo.  Dorsal  view,  x 1. 

Fig.  9.  Eccoptochile  ( Eccoptochile ) cf.  clavigera  (Beyrich).  External  impression.  SG  (no  number,  in  drawer 
labelled  3A2).  1400  m S 32°  E of  Covelo  church,  north  Portugal.  Upper  part  of  Valongo  Formation;  Lower 
Llandeilo.  Dorsal  view,  x 1-5. 


PLATE  79 


romano,  Ordovician  trilobite  Eccoptochile 


612 


PALAEONTOLOGY,  VOLUME  23 


Llandeilo:  SMF  24784  from  Navatrasierra  (Montes  de  Toledo,  Spain);  Na  la.  Lower  Llandeilo:  SMF  24784 
from  Navatrasierra  (Montes  de  Toledo,  Spain);  Na  la.  Lower  Llandeilo:  SMF  24787  from  Navatrasierra 
(Montes  de  Toledo);  Na  la,  basal  Llandeilo. 

Description.  The  types  from  Spain  and  Portugal  have  been  well  described  and  figured  by  Hammann  (1974) 
and  Curtis  (1961).  No  further  comments  are  necessary. 

Discussion.  Verneuil  and  Barrande  erected  Eccoptochile  (l  Eccoptochile)  mariana  (1855,  p.  970,  pi.  23, 
fig.  4)  on  the  basis  of  it  having  a more  dorsally  convex  glabella  and  a wider,  flat  anterior  border  than 
Eccoptochile  ( Eccoptochile ) clavigera  (Beyrich,  1845).  They  stated  that  the  eye  occupied  the  same 
relative  position  in  both  species.  Curtis  (1961,  p.  8)  listed  four  differences  between  the  two  species, 
including  that  in  E.  (IE.)  mariana  ( sensu  Curtis  and  Hammann)  the  frontal  lobe  is  relatively  shorter, 
the  eye  ridge  starts  level  with  the  anterior  glabella  furrow  and  the  eye  is  situated  farther  back.  The 
specimen  figured  by  Curtis  (1961,  pi.  1,  fig.  1)  as  E.  mariana , and  later  referred  to  E.  cf.  mariana 
by  Hammann  (1974,  p.  105)  possesses  a wide  frontal  area  which  distinguishes  it  from  other  specimens 
of  E.  (IE.)  mariana  as  understood  by  Curtis  and  Hammann.  A reinvestigation  of  the  holotype  of 
E.  (IE.)  mariana  also  revealed  the  presence  of  a wide  frontal  area  and  it  is  thus  clearly  distinct 
from  the  majority  of  specimens  previously  assigned  to  that  species.  The  evenly  curved  longitudinal 
profile  of  the  glabella  of  the  holotype  (text-fig.  2k  and  PI.  78,  fig.  3)  is  also  unlike  that  in 
E.  (IE.)  mariana  sensu  Hammann  where  maximum  curvature  occurs  in  the  anterior  part  of  the 
glabella  (Hammann  1974,  pi.  12,  fig.  1926).  Thus  E.  (IE.)  mariana  is  restricted  in  this  paper  to 
include,  with  the  holotype,  only  the  two  specimens  from  the  Valongo  area;  that  figured  by  Curtis 
(1961,  pi.  1,  fig.  1 and  refigured  here,  PI.  78,  figs.  5,  6)  and  a previously  unfigured  specimen 
(PI.  78,  fig.  7). 

The  relative  lengths  (sag.)  of  the  frontal  glabellar  lobe  and  frontal  area  appear  to  show  significant 
differences  in  the  species  almadenensis,  clavigera , and  mariana.  In  an  attempt  to  quantify  these 
differences  the  three  parameters  B,  F,  and  b5  (text-fig.  3a)  (symbols  from  Shaw,  1957  and  Temple, 


text-fig.  3.  a.  Outline  of  cranidium  of  Eccoptochile 
(Eccoptochile)  almadenensis  sp.  nov.  (after  Ham- 
mann, 1974,  text-fig.  39)  showing  parameters  used  in 
(b),  (c)  and  text-fig.  4;  b,  c.  Scatter  diagrams  of  F 
against  B and  F against  b5  respectively  with  calculated 
regression  lines  for  the  species  almadensis,  clavigera 
and  mariana. 


ROMANO:  ORDOVICIAN  TRILOBITE  ECCOPTOCHILE 


613 


1975)  were  selected  since  it  is  assumed  the  ratio  of  these  measurements  taken  along  a constant 
orientation  will  be  virtually  unaffected  by  deformation.  When  the  three  parameters  are  plotted  on 
size  frequency  and  scatter  diagrams  the  species  plot  out  in  isolated  and  relatively  restricted  fields. 
Size/frequency  histograms  of  the  B:F  and  B : b5  ratios  (not  illustrated)  serve  to  distinguish 
E.  (IE.)  mariana  from  E.  (E.)  almadenensis  and  E.  ( E .)  clavigera  quite  markedly.  The  regression 
lines  of  B against  F and  b5  against  F (text-fig.  3b  and  3c)  show  that  for  mariana  at  least  the  lines 
appear  to  be  clearly  distinguishable  and  although  few  specimens  were  available  to  construct  the 
graphs  (almadenensis— 14;  clavigera— 7;  mariana— 3)  the  contrast  in  gradient  suggests  the  difference 
in  growth  rate  is  a useful  criterion  for  separating  this  species.  When  the  three  parameters  are  plotted 
as  ratios  on  a triangular  graph  (text-fig.  4)  the  three  species  plot  out  in  discrete  fields  and  the 


B 


text-fig.  4.  Triangular  plot  for  the  species  almadenensis,  clavigera,  and  mariana 
using  the  three  parameters  B,  F,  b5  (see  text-fig.  3).  For  material  and  references 
used  to  construct  the  graph  see  text.  Additional  sources  include  Dr.  J.-L.  Henry 
(pers.  comm.)  and  author’s  collection,  University  of  Sheffield. 


selected  holotype  for  E.  (E.)  almadenensis  occurs  near  the  centre  of  scatter  for  that  species.  Since 
the  number  of  specimens  is  small  the  fields  have  not  been  numerically  defined.  The  species 
clavigera  is  clearly  distinguishable  by  the  presence  of  a long  (sag.)  frontal  glabellar  lobe  (see  text-fig.  2) 
and  the  flat  profile  of  the  glabella  in  lateral  view.  This  difference  in  the  relative  length  of  the  frontal 
lobe  is  shown  in  the  groupings  in  text-fig.  4. 

Using  the  methods  outlined  above,  E.  (IE.)  guillieri  cannot  be  distinguished  from  E.  (E.) 
almadenensis  since  measurements  taken  from  the  photographs  supplied  by  Dr.  J.-L.  Henry  plot  out 
near  the  middle  of  the  E.  (E.)  almadenensis  field  (text-fig.  4).  However,  the  strong  glabella  convexity 


614 


PALAEONTOLOGY,  VOLUME  23 


and  subrounded  outline  of  the  glabella  in  dorsal  view  of  E.  (IE.)  guillieri  are  characteristic 
enough  to  suggest  it  is  a valid  species.  The  specimen  listed  by  Delgado  (1908,  p.  106),  as 
‘ Cheirurus  sp.  n.  (aff.  Ch.  Sedgwicki  McCoy)’,  from  1400  m S 32°  E of  Covelo  church  (SG  1711) 
appears  to  show  no  important  differences  from  E.  (E.)  almadenensis.  The  size  of  the  free  cheek, 
position,  and  structure  of  the  eye  in  Placoparina  sedgwicki  (Whittard,  1958,  pp.  112,  115)  are 
distinctive,  and  although  the  Portuguese  specimen  listed  by  Delgado  is  imperfectly  preserved 
(PI.  79,  fig.  5)  it  is  assigned  to  E.  (E.)  almadenensis.  Delgado  (1908,  p.  106)  also  recorded 
‘ Cheirurus  Guillieri  Trom.  (aff.  Ch.  claviger  Beyr.)’  from  the  Valongo  area,  1650  m S 20°  W from 
the  hill  of  Santa  Justa  (SG  1704),  but  the  forwardly  expanding  and  relatively  longer  glabella  (PI.  79, 
fig.  1)  is  unlike  that  of  the  holotype  of  E.  (IE.)  guillieri  and  this  specimen  is  also  identified  as 
E.  (E.)  almadenensis.  Another  specimen  (PI.  79,  fig.  8)  identified  by  Delgado  (op.  cit.)  as  ‘ Cheirurus 
Guillieri ’ is  here  referred  to  E.  (IE.)  cf.  mariana  because,  although  it  closely  resembles  the  holotype, 
the  deformed  specimen  precludes  a definite  identification.  The  eccoptochilinid  from  Perhaver  Beach, 
Cornwall,  tentatively  identified  as  E.  (E.)  clavigera  by  Whittard  (1958,  p.  115)  and  Sadler  (1974, 
p.  73)  is  an  incomplete  cranidium  (PI.  78,  fig.  8)  which  can  now  be  confidently  assigned  to  E.  (E.) 
almadenensis. 


Eccoptochile  ( Eccoptochile ) cf.  clavigera  (Beyrich,  1845) 

Plate  79,  fig.  9 

Figured  material.  One  external  impression  of  an  incomplete  flattened  pygidium;  specimen  housed  in  Serv^os 
Geologicos,  Lisbon;  drawer  3A2. 

Horizon  and  locality.  1400  m S 32°  E of  Covelo  church,  Valongo;  probably  from  upper  part  of  Valongo 
Formation,  probably  Lower  Llandeilo. 

Description.  Pygidium  nearly  twice  as  wide  as  long.  Anterior  margin  gently  rounded  with  nearly  straight 
median  portion  and  more  strongly  rounded  posterior  margin.  Axis  subtriangular  in  outline  (articulating  half 
ring  not  preserved)  with  outwardly  curved  axial  furrows.  Axis  probably  slightly  wider  than  long,  reaching 
back  to  about  one-half  length  of  pygidium;  three  axial  rings  and  a small  triangular  terminal  piece;  rings  decrease 
in  length  posteriorly,  ring  furrows  shallow  medially  (except  third  axial  ring  furrow).  Axial  furrows  shallow  and 
weakly  defined  and  not  present  posterior  to  the  second  axial  ring  furrow.  Three  pairs  of  broad,  bluntly 
rounded,  spinose  pleural  ribs.  First  and  second  ribs  start  opposite  first  two  axial  rings  and  curve  gently 
outwards  and  backwards;  third  pair  directed  posteriorly.  7-8  shallow  pits  on  first  pleural  ribs  situated  at  about 
midlength  (exsag.)  of  rib  and  extend  for  about  one-quarter  along  the  rib.  Only  1-2  pits  are  present  on  the 
second  rib  and  none  on  the  third.  Surface  of  pygidium  covered  with  fine,  closely  spaced  tubercles  except  in  the 
shallow  rib  pits. 

Discussion.  The  poor  preservation  of  this  specimen  makes  it  difficult  to  compare  length  to  width 
ratios  with  the  type  material  of  E.  (E.)  clavigera  (Beyrich,  1845,  plate  (unnumbered),  fig.  3),  which 
appears  to  be  relatively  wider.  In  all  other  respects  it  closely  resembles  the  holotype.  The  present 
material  is  very  similar  to  the  specimen  referred  to  E.  (E.)  clavigera  by  Pribyl  and  Vanek  (1969, 
p.  3,  fig.  8)  except  that  in  the  latter  the  rows  of  pits  on  the  pleural  ribs  extend  further  along  the 
rib,  although  this  is  not  so  apparent  in  other  specimens  figured  by  those  authors  (op.  cit.  pi.  3, 
figs.  6,  7). 


RANGE  AND  DISTRIBUTION  OF  E.  ( E .)  ALMADENENSIS,  E.  (E.)  CLAVIGERA, 
AND  E.  (IE.)  MARIANA 

E.  (E.)  almadenensis  is  the  most  widespread  species  in  Iberia  and  the  Armorican  Massif  and 
probably  also  occurs  in  southern  Cornwall.  It  first  appears  in  the  basal  Llandeilo  of  Navatrasierra 
in  central  Spain  (Hamman  1974,  p.  15)  and  occurs  in  the  Lower  Llandeilo  of  north  Portugal,  the 
Armorican  Massif,  and  probably  southern  England.  There  is  evidence  that  the  species  possibly  also 
persists  into  the  Caradoc  in  the  region  south  of  Rennes,  Brittany  (Lindstrom  et  al.,  1974,  p.  20).  There 


ROMANO:  ORDOVICIAN  TRILOBITE  ECCOPTOCHILE 


615 


is  no  record  of  it  continuing  into  the  Ashgill.  E.  (IE.)  mariana  (as  understood  in  this  paper)  is  a 
relatively  restricted  species,  recorded  only  from  the  Ciudad  Real  region  in  south  central  Spain  where  it 
is  of  Upper  Llandeilo  age  and  from  the  area  around  Covelo,  near  Valongo  in  north  Portugal 
(Lower  Llandeilo).  E.  ( E .)  clavigera  is  poorly  represented  in  Spain  and  north  Portugal;  E.  (E.) 
cf.  clavigera  (a  deformed  pygidium)  occurs  in  probably  Lower  Llandeilo  beds  in  the  Valongo  area 
and  E.  ( E .)  aff.  clavigera  (a  hypostoma  and  pygidium)  is  recorded  from  Caradoc  beds  north  of 
Almaden,  central  Spain  (Hammann  1974,  p.  1 1 1).  A deformed  eccoptochilinid  cranidium  from  the 
?Caradoc  of  central  Portugal,  50  km  SSE  of  Coimbra  (A.  H.  Cooper  collection),  is  probably  referable 
to  E.  ( E .)  clavigera  and  the  Delgado  collection  housed  in  the  Servigos  Geologicos,  Lisbon,  contains 
large  specimens  of  E.  (E.)  clavigera  from  the  Magao  region  80  km  SSE  of  Coimbra.  The  age  of  the 
Magao  specimens  is  not  known  but  the  associated  fauna  contain  Actinopeltis  and  Eoharpes  and  could 
indicate  an  Upper  Llandeilo  to  Caradoc  age.  E.  (E.)  clavigera  is  common  in  Bohemia  where  it 
ranges  from  the  Liben  Formation  to  the  Bohdalec  Formation  (Havlicek  and  Vanek  1966)  and  is 
associated  with  Actinopeltis.  Havlicek  and  Marek  (1973)  have  revised  the  chronostratigraphic 
terminology  for  the  Bohemian  sequence  and  they  recognize  a Beroun  Series  of  middle  Llandeilo  to 
upper  Caradoc  age  which  includes  the  range  of  E.  (E.)  clavigera.  In  Bohemia  Eoharpes  dies  out  in  the 
Dobrotiva  Formation  which  is  considered  by  these  authors  to  be  equivalent  in  age  to  the  lower  part 
of  the  Llandeilo. 

Any  conclusions  regarding  faunal  migrations  and  phylogeny  within  the  group  must  await  further 
work  in  particular  on  the  existing  collections  in  Lisbon. 

Acknowledgements.  I thank  Dr.  R.  A.  Fortey  (British  Museum),  Dr.  A.  W.  A.  Rushton  (Institute  of  Geological 
Sciences),  and  the  Director  of  the  Ecole  National  Superieure  des  Mines  for  loaning  material  in  their  care. 
I also  thank  Dr.  Jean-Louis  Henry  for  supplying  me  with  photographs  and  outline  drawings  of  eccoptochilinid 
trilobites  from  the  Armorican  Massif.  He  and  Professor  H.  B.  Whittington  kindly  read  and  criticized  the 
manuscript.  Mr.  M.  Cooper  redrew  the  diagrams  and  Miss  P.  Mellor  typed  the  manuscript.  The  work  was  made 
possible  by  a N.E.R.C.  grant. 


REFERENCES 

barrande,  j.  1852.  Systeme  silurien  du  centre  de  la  Boheme.  lere  partie.  Recherches  paleontologiques,  vol.  1. 

Crustaces,  Trilobites.  xxx  + 935  pp.,  51  pis.  Prague  and  Paris. 
barton,  d.  c.  1916.  A revision  of  the  Cheirurinae  with  notes  on  their  evolution.  Wash.  Univ.  Stud,  scient.  Ser. 
3(1),  101-152. 

beyrich,  E.  1845.  Ueber  einige  bohmische  Trilobiten.  47  pp.,  1 pi.  Berlin. 

born,  a.  1916.  Die  Calymene  Tristani- Stufe  (mittleres  Untersilur)  bei  Almaden,  ihre  Fauna,  Gliederung  und 
Verbreitung.  Abhandl.  Senck.  Naturf.  gesell.  36,  309-358,  4 pis. 
costa,  j.  s.  Carrington  da,  1931.  O Paleozdico  Portugues  ( Sintese  e Critica),  1-141,  Porto. 

Curtis,  m.  l.  K.  1961.  Ordovician  trilobites  from  the  Valongo  area,  Portugal.  Cheiruridae,  Pliomeridae,  and 
Dionididae.  Bol.  Soc.  Geol.  Port,  xiv,  1-16,  4 pis. 

delgado,  J.  F.  n.  1892.  Fauna  Silurica  de  Portugal.  Descripgao  de  uma  forma  nova  de  trilobite — Lichas 
(Uralichas)  Ribeiroi.  (With  French  translation.)  Comm.  Trab.  Geol.  Port.  Lisbon,  1-32,  6 pis. 

— 1897.  Fauna  Silurica  de  Portugal.  Novas  observagoes  acerca  de  Lichas  ( Uralichas ) Ribeiroi.  (With  French 
translation.)  Dir.  Trab.  Geol.  Port.  Lisbon,  1-34,  4 pis. 

— 1908.  Systeme  silurique  du  Portugal;  etude  de  stratigraphie  paleontologie.  Mem.  Commn  geol.  Port. 
Lisbon,  1-245,  8 pis. 

guillier,  a.  1873.  Faune  seconde  silurienne  entre  Saint-Denis-d’Orques  et  Chemire-en-Charnie.  Bull.  Soc. 
Agric.  Sci.  Arts  Sarthe,  21,  633-636. 

hammann,  w.  1971a.  Die  Placopariinae  (Trilobita,  Cheirurina;  Ordovizium).  Senck.  Lethaea,  52,  53-75,  3 pis. 
— 1971  b.  Stratigraphische  Einteilung  des  spanischen  Ordoviziums  nach  Dalmanitacea  und  Cheirurina 
(Trilobita).  Colloque  Ordovicien-Silurien  Brest  1971.  Mem.  Bur.  Rech.  geol.  minier.  73,  265-272,  1 pi. 

— 1974.  Phacopina  und  Cheirurina  (Trilobita)  aus  dem  Ordovizium  von  Spanien.  Senck.  Lethaea.  55,  1-150, 
12  pis. 

Harrington,  H.  j.  1959.  In  moore,  R.  c.  (Editor).  Treatise  on  Invertebrate  Paleontology.  Part  O,  Arthropoda  1. 
i-xix,  560  pp.  Geol.  Soc.  Amer.  and  Univ.  Kansas  Press. 


616 


PALAEONTOLOGY,  VOLUME  23 


havlicek,  v.  and  marek,  L.  1973.  Bohemian  Ordovician  and  its  international  correlation.  Cas.  Miner,  geol.  18, 
225-232. 

— and  vanek,  J.  1966.  The  biostratigraphy  of  the  Ordovician  of  Bohemia.  Shorn,  geol.  ved.,  pal.  8, 
7-68,  16  pis.  (Czech  summary,  p.  69.) 

henningsmoen,  G.  1959.  In  moore,  R.  c.  (Editor).  Treatise  on  Invertebrate  Paleontology.  Part  O,  Arthropoda  1. 

i-xix,  560  pp.  Geol.  Soc.  Amer.  and  Univ.  Kansas  Press. 
henry,  j.-l.  and  clarkson,  e.  n.  k.  1975.  Enrollment  and  coaptations  in  some  species  of  the  Ordovician 
trilobite  genus  Placoparia.  Fossils  Strata,  4,  87-95,  3 pis. 

and  romano,  m.  1978.  Le  genre  Dionide  Barrande,  1847  (Trilobite)  dans  l’Ordovicien  du  Massif  Armoricain 

et  du  Portugal.  Geobios,  11,  327-343,  2 pis. 

lane,  p.  d.  1971.  British  Cheiruridae  (Trilobita).  Palaeontogr.  Soc.  ( Monogr .),  95  pp.,  16  pis. 
lindstrom,  m.,  rachebouef,  p.  r.  and  henry,  j.-l.  1974.  Ordovician  conodonts  from  the  Postolonnec  Forma- 
tion (Crozon  peninsula,  Massif  Armoricain)  and  their  stratigraphic  significance.  Geol.  et  Palaeont.  8, 
15-23,  2 pis. 

prantl,  f.  and  pribyl,  a.  1948.  Rostrideni  nekterych  ceskych  Cheiruridu.  (Trilobitae).  (Classification  of  some 
Bohemian  Cheiruridae.)  Sb.  nar.  Mus.  Praze,  ( B ) Geol.  ( Paleont .),  1,  1-44,  6 pis. 
pribyl,  A.  and  vanek,  j.  1969.  Uber  einige  Trilobiten  des  mittelbohmischen  Ordoviziums.  Vestnik.  Ustr.  ust. 
geol.  44,  365-374,  6 pis. 

rachebouef,  p.  r.  1969.  Generalites  sur  quelques  trilobites  des  schistes  Ordoviciens  de  la  Mayenne.  Bull. 
Bayenne-Sci.,  66-86,  6 pis. 

reed,  f.  r.  c.  1896.  Notes  on  the  evolution  of  the  genus  Cheirurus.  Geol.  Mag.  (4)  3,  117-123,  161-167. 
reid,  c.  1907.  Explanation  of  Sheet  353.  The  geology  of  the  country  around  Mevagissey.  Mem.  geol.  Surv.  Eng. 
Wales,  vi  + 73  pp.  7 pis. 

romano,  M.  1975.  Harpid  trilobites  from  the  Ordovician  of  North  Portugal.  Comm.  Serv.  Geol.  Port.  59, 
27-36,  1 pi. 

— 1976.  The  trilobite  genus  Placoparia  from  the  Ordovician  of  the  Valongo  area,  north  Portugal.  Geol. 
Mag.  113(1),  11-28,  1 pi. 

and  diggens,  J.  N.  1973-1974.  The  stratigraphy  and  structure  of  Ordovician  and  associated  rocks  around 

Valongo,  north  Portugal.  Com.  Serv.  Geol.  Port.  (edit.  1976),  57,  23-50,  2 pis. 
sadler,  p.  m.  1974.  Trilobites  from  the  Gorran  Quartzites,  Ordovician  of  south  Cornwall.  Palaeontology,  17, 
71-93,  pis.  9,  10. 

Salter,  J.  w.  1853.  Notes  on  the  trilobites.  (Appendix  C to  ‘On  the  Carboniferous  and  Silurian  formations  of 
the  neighbourhood  of  Bussaco  in  Portugal’  by  C.  Ribeiro.)  Quart.  J.  Geol.  Soc.  9,  158-60,  3 pis. 
sharpe,  d.  1849.  On  the  geology  of  the  neighbourhood  of  Oporto,  including  the  Silurian  coal  and  slates  of 
Vallongo.  Ibid.  5,  142-153. 

shaw,  A.  b.  1957.  Quantitative  trilobite  studies.  II.  Measurement  of  the  dorsal  shell  of  non-agnostidean 
trilobites.  J.  Paleont.  31,  193-207. 

teixeira,  c.  1955.  Not  as  sobre  geologia  de  Portugal.  O Sistema  Silurico.  Lisboa. 

temple,  j.  t.  1975.  Standardization  of  trilobite  orientation  and  measurement.  Fossils  Strata,  4,  461-467. 
thadeu,  D.  1947.  Trilobites  do  silurico  de  Loredo  (Bu5aco).  Bol.  Soc.  Geol.  Port.  6,  217-236,  3 pis. 

1956.  Note  sur  le  Silurien  Beiro-Durien.  Ibid.  12,  1-38,  9 pis. 

tromelin,  G.  and  lebesconte,  p.  1876.  Essai  d’un  catalogue  raisonne  des  fossiles  silurien  des  departements  de 
Maine-et-Loire,  de  la  Loire-Inferieure  et  du  Morbihan,  avec  des  observations  sur  les  terrains  paleozoiques 
de  l’ouest  de  la  France.  C.R.  4e  Congr.  Assoc,  franc.  Avancem.  Sci.  601-661. 
verneuil,  p.  e.  p.  and  barrande,  J.  1855.  Descriptions  des  fossiles  trouves  dans  les  terrains  silurien  et  devonien 
d’Almaden,  d’une  partie  de  la  Sierra  Morena  et  des  montagnes  de  Tolede.  Bull.  Soc.  Geol.  France,  12, 


whittard,  w.  f.  1958.  The  Ordovician  trilobites  of  the  Shelve  Inlier,  West  Shropshire.  Palaeontogr.  Soc. 
{Monogr.),  71-116,  pis.  x-xv. 


904-1025. 


Typescript  received  13  June  1979 

Revised  typescript  received  22  November  1979 


M.  ROMANO 

Department  of  Geology 
Beaumont  Building 
University  of  Sheffield 
Sheffield  S3  7HF 


THE  MIOCENE  HORSE  HIPPARION  FROM 
NORTH  AMERICA  AND  FROM  THE  TYPE 
LOCALITY  IN  SOUTHERN  FRANCE 

by  BRUCE  J.  MACFADDEN 


Abstract.  The  three-toed  horse  Hipparion  is  diagnosed  by  the  presence  of  a preorbital  facial  fossa  that 
anteriorly  is  poorly  defined  and  posteriorly  is  moderately  pocketed  with  a well-developed  and  continuous  rim. 
The  concept  of  the  genus  Hipparion  sensu  s trie  to  ( s.s .)  is  presently  restricted  in  the  Old  World  to  H.  prostylum 
from  the  genotypic  locality  at  Mt.  Leberon,  France,  and  the  species  H.  tehonense  and  H.forcei  from  New  World 
localities  with  a similar  configuration  of  the  preorbital  facial  fossa.  It  has  previously  been  stated  that,  although 
Hipparion  was  common  in  the  Old  World  Neogene,  this  genus  was  very  rare  in  equivalent-aged  sediments  in  the 
New  World.  Based  on  the  concept  of  the  genus  presented  here,  Hipparion  s.s.  is  found  at  numerous  New  World 
localities.  There  apparently  was  a generic-level  continuity  of  Hipparion  s.s.  that  existed  throughout  Holarctica 
during  part  of  the  Neogene.  Hipparion  horses  ( sensu  lato)  appear  to  represent  a polyphyletic  assemblage  of 
several  genera  that  arose  independently  from  more  than  one  merychippine  ancestor  during  the  Miocene.  The 
presence  of  hipparion  horses  in  the  New  and  Old  Worlds  probably  resulted  from  more  than  one  dispersal 
event  across  Beringia. 

For  more  than  a century,  the  genus  Hipparion  has  been  used  as  a horizontal  taxon,  or  ‘form  genus’, 
to  include  Holarctic  Mio-Pliocene  horses  with  isolated  protocones  in  the  upper  molars,  and  tridactyl 
limbs.  The  great  geographic  and  geological  abundance  of  this  horse  has  made  it  biostratigraphically 
very  useful  for  Neogene  intercontinental  correlations.  More  than  one  hundred  species  of  ‘Hipparion' 
{sensu  lato)  have  been  named  primarily  on  dental  and  postcranial  characters.  This  large  complex 
of  species  is  so  unwieldy  that,  rather  than  comparing  a new  sample  to  all  the  existing  species, 
palaeontologists  often  propose  new  species  out  of  despair  and  therefore  perpetuate  this  taxonomic 
problem. 

In  recent  years,  several  studies  have  been  presented  that  attempt  to  sort  out  some  of  the  different 
hipparion  groups  based  principally  on  cranial  morphology.  Skinner  and  MacFadden  ( 1 977)  analysed 
relatively  large  quarry  samples  from  the  North  American  mid-continent  and  showed  that  the 
development  of  the  preorbital  facial  fossa  appears  to  be  a taxonomically  valid  character  complex  at 
the  generic  rank.  In  their  study  they  proposed  the  genus  Cormohipparion  for  hipparions  with  a 
diagnostic  preorbital  (also  termed  nasomaxillary)  facial  fossa  that  is  pocketed  posteriorly  and  has 
well-developed  and  continuous  anterior  and  posterior  rims.  Skinner  and  MacFadden  (1977)  con- 
centrated mostly  on  North  American  forms  but  also  tentatively  referred  some  Eurasian  hipparions 
to  this  genus.  MacFadden  and  Bakr  (1979)  studied  the  Siwalik  hipparions  from  the  Indo-Pakistan 
subcontinent  and  refer  the  large  species  theobaldi  to  the  genus  Cormohipparion.  Woodburne  and 
Bernor  (1980)  studied  numerous  museum  collections  of  Eurasian  hipparions  and  proposed  several 
distinct  groups,  which  probably  represent  separate  lineages,  based  principally  on  their  analysis  of 
cranial  characters.  There  is  general  agreement  among  students  of  equid  systematics  that  one  or  more 
members  of  this  polyphyletic  hipparion  assemblage  arose  in  North  America  during  the  medial 
Miocene.  Subsequently,  it  appears  that  more  than  one  hipparion  group  (i.e.  a few  genera)  dispersed 
into  the  Old  World  during  the  later  Miocene.  Many  workers  have  suggested  that  the  presence  of 
hipparions  in  the  Old  World  resulted  from  the  dispersal  of  one  monophyletic  group  or  ‘species’  of 
‘ Hipparion ’ (e.g.  Forsten  1968;  Hussain  1971).  Skinner  and  MacFadden  (1977)  suggested,  based  on 
different  cranial  morphologies,  that  the  dispersal  of  hipparions  from  the  New  to  the  Old  World  was 
not  monophyletic  and  probably  involved  several  forms  (or  genera). 

[Palaeontology,  Vol.  23,  Part  3,  1980,  pp.  617-635.| 


618 


PALAEONTOLOGY,  VOLUME  23 


The  concept  of  the  genus  Hipparion  sensu  stricto  ( s.s .)  is  based  on  the  species  H.  prostylum 
described  from  the  Turolian  Mt.  Leberon  locality  in  southern  France  (de  Christol  1832).  One  of  the 
important  problems  in  the  study  of  hipparion  systematics  has  been  recognition  of  the  genus 
Hipparion  s.s.  in  North  America.  Gidley  (1903)  proposed  the  genus  Neohipparion  for  most  of  the  New 
World  species  that  had  been  previously  included  in  the  genus  Hipparion,  and  Hipparion  s.s.  was 
almost  exclusively  used  for  Old  World  forms.  Osborn  (1918)  did  not  strictly  follow  Gidley’s 
dichotomy  between  Neohipparion  and  Hipparion  for  New  versus  Old  World  forms,  respectively.  Since 
the  early  studies,  many  workers  believed  that  Hipparion  s.s.  was  abundant  in  the  Old  World  Miocene 
and  rare  in  the  New  World.  Stirton  (1940)  stated  that  in  North  America  Hipparion  s.s.  was 
represented  by  only  a few  species  distributed  in  California,  Oregon,  Washington,  and  Florida. 

The  purpose  of  this  report  is  to  describe  Hipparion  sensu  stricto  from  several  localities  in  North 
America  and  to  compare  these  samples  with  the  material  from  the  genotypic  locality  in  southern 
France.  This  study  shows  that  Hipparion  s.s.  was  more  widely  distributed  in  North  America  than  has 
been  previously  thought.  Only  the  North  American  localities  with  well-preserved  cranial  material  are 
discussed  here.  Hipparion  s.s.  is  undoubtedly  present  at  numerous  other  localities  in  North  America, 
however,  without  relevant  cranial  material,  it  is  difficult  to  distinguish  these  occurrences.  It  is  not  the 
purpose  of  this  paper  to  revise  the  taxonomy  of  all  species  of  Hipparion  and  related  forms,  as  that  task 
would  certainly  require  a monograph.  Therefore,  the  specific  diagnoses  and  assignments  essentially 
rely  on  previous  studies.  The  phylogenetic  and  palaeogeographic  implications  presented  at  the  end  of 
the  present  study  will  focus  on  the  recognition  of  a generic-level  continuity  of  Hipparion  s.s.  through- 
out Holarctica  during  the  late  Miocene. 

The  following  institutional  abbreviations  are  used  in  the  text:  AMNH,  Department  of  Vertebrate 
Paleontology,  American  Museum  of  Natural  History,  New  York;  BMNH,  Department  of  Palaeonto- 
logy, British  Museum  (Natural  History),  London;  CIT,  California  Institute  of  Technology  Collec- 
tion, now  housed  at  the  Los  Angeles  County  Museum  of  Natural  History,  Los  Angeles;  F:AM,  Frick 
American  Mammals,  The  American  Museum  of  Natural  History,  New  York;  MNHNP,  Museum 
National  d’Histoire  Naturelle,  Institut  de  Paleontologie,  8 rue  de  Buffon,  Paris  5,  France;  UCMP, 
University  of  California  Museum  of  Paleontology,  Berkeley;  UF,  Florida  State  Museum,  University 
of  Florida,  Gainesville.  The  dental  nomenclature  follows  Stirton  (1940,  1941),  Skinner  and  Taylor 
(1967),  and  Skinner  and  MacFadden  (1977). 

SYSTEMATIC  PALAEONTOLOGY 

Class  mammalia  Linnaeus,  1758 
Order  perissodactyla  Owen,  1848 
Family  equidae  Gray,  1821 
Genus  hipparion  de  Christol,  1832 

Text-figs.  1-14 

Type  status.  When  de  Christol  (1832)  first  proposed  the  genus  Hipparion  based  on  material  from  Mt.  Leberon 
in  southern  France  (also  called  Mt.  Luberon,  Cucuron),  no  holotype  was  indicated.  Later,  Gervais  (1849) 
designated  a syntypic  series  of  Hipparion  from  Mt.  Leberon,  including  H.  prostylum,  H.  mesostylum,  and 
H.  diplostylum.  Osborn  (1918)  considered  H.  prostylum  to  be  the  type  species  for  the  genus  Hipparion.  Sondaar 
(1974)  stated  that  the  holotype  of  H.  prostylum,  which  consists  of  a fragmentary  palate  with  P4-M2  (see  Gervais 
1849,  pi.  19,  fig.  2),  is  probably  contained  in  the  collections  in  the  Musee  Requien,  Avignon. 

Revised  generic  diagnosis.  Medium-sized,  mesocephalic,  and  moderately  hypsodont  tridactyl  horses.  Nasal 
notch  moderately  developed  and  extends  posteriorly  to  a position  anterior  to,  or  lying  over,  P2.  Infra- 
orbital foramen  lies  over  P3.  Preorbital  facial  fossa  lies  dorsal  to  P3-M*  on  the  nasal  and  maxillary  bones 
well  forward  of  the  anterior  rim  of  the  orbit.  The  posterior  portion  of  the  fossa  is  usually  developed  on  the 
nasal  and  maxillary  bones,  anterior  to  the  lacrimal.  Anteriorly  the  fossa  is  poorly  defined  and  is  confluent  with 
the  facial  region.  Posteriorly  the  fossa  is  moderately  pocketed  and  has  a well-developed  and  continuous  rim. 
There  is  no  ventral  fossa  associated  with  the  malar  crest  as  is  the  case  in  some  other  horses.  In  the  upper 
cheek  teeth  the  protocones  vary  from  rounded  to  oval  to  lenticulate.  There  is  a tendency  for  the  protocone  to 


M ACFADDEN:  MIOCENE  HORSE  HIPPARION 


619 


be  connected  to  the  protoloph  in  earlier  wear  stages  than  some  other  hipparions,  e.g.  Neohipparion.  The  hypo- 
conal  groove  is  moderately  developed  and  is  distinct  to  the  base  of  the  tooth.  In  the  lower  cheek  teeth  there  is  a 
progressive  deepening  of  the  ectoflexids  posteriorly.  The  metaconids  and  metastylids  are  widely  separated.  The 
parastylid  (also  termed  ectoparastylid  or  protostylid)  is  often  developed  and  is  either  connected  to  the  proto- 
conid  or  is  isolated.  In  both  the  upper  and  lower  cheek  teeth  the  enamel  plications  vary  from  simple  to 
moderately  developed. 

Distribution.  Late  Miocene  (Clarendonian-?early  Hemphillian)  of  North  America,  late  Miocene-?Pliocene 
(Vallesian-?Villafranchian)  of  Eurasia,  and  possibly  Miocene-Pliocene  of  Africa.  Note.  The  questionable  ranges 
listed  here  are  taken  from  previous  studies  in  which  relevant  cranial  material  is  lacking.  Therefore,  it  is  difficult 
to  allocate  certain  Old  World  species  to  the  genus  Hipparion  s.s. 

Included  species.  At  this  point  it  is  impossible  to  list  all  the  species  that  should  be  included  in  Hipparion  s.s. 
(particularly  in  the  Old  World)  because  of  the  problems  in  recognition  of  this  genus  without  cranial  material.  In 
the  present  report  H.  tehonense  and  H.  forcei  are  described  from  North  America  and  these  are  compared  to 
H.  prostylum  from  Europe. 


Hipparion  prostylum  Gervais,  1849 
Text-figs.  1-5,  13,  14 


Selected  synonymy 

1849  Hipparion  prostylum  Gervais,  pp.  284-285. 

1873  Hipparion  gracile  Gaudry,  pp.  32-42,  pi.  5,  figs.  7-10;  pi.  6,  figs.  1-11;  pi.  7,  fig.  1. 

1956  Hipparion  mediterraneum  (in  part),  Pirlot,  p.  28. 

1968  Hipparion  mediterraneus  (in  part),  Forsten,  pp.  40-53,  83-129,  tables  12-15. 

1974  Hipparion  prostylum  Sondaar,  pp.  289-290,  296-299,  301-306,  tables  2-4,  pi.  46,  figs.  1,  2;  pi.  48, 
figs.  2,  3,  8-10;  pi.  49,  figs.  3,  4,  8,  9,  10. 

Type  specimen.  See  generic  discussion. 

Specific  diagnosis.  Same  as  for  the  genus  with  the  limitation  that  H.  prostylum  has  rounded  (and  infrequently 
oval)  protocones  in  the  upper  molars.  Sondaar  (1974,  p.  297,  adapted  from  Gromova  1952)  diagnoses 
H.  prostylum  as  follows:  ‘Average  size,  length  of  the  upper  molar  series  P2-M2  123-145  mm.  Enamel  with 
little  foldings,  slender  footbones  with  relatively  long  metapodials.’  See  discussion  below. 

Referred  material.  This  description  is  based  on  the  collections  of  H.  prostylum  housed  in  Paris  (NMNHP)  and 
London  (BMNH).  These  collections  consist  of  four  skulls,  numerous  dentitions,  isolated  teeth,  and  postcranials. 

Distribution.  Hipparion  prostylum  is  recognized  at  the  type  locality,  Mt.  Leberon,  which  is  of  Turolian  (late 
Miocene)  age.  This  species  is  also  part  of  the  ‘hipparionine  Group  3’  complex  of  Woodburne  and  Bernor  (1980). 
Therefore,  H.  prostylum  is  probably  represented  at  several  other  Old  World  localities  of  Turolian  age  listed  in 
that  publication.  Pending  a revision  of  hipparions  from  other  Old  World  localities,  H.  prostylum  is  presently 
only  known  to  occur  for  certain  at  the  type  locality,  Mt.  Leberon. 

Description.  Although  H.  prostylum  has  been  described  elsewhere  (e.g.  Gaudry  1873;  Gromova  1952;  Sondaar 
1974)  it  is  redescribed  in  this  report  in  order  to  compare  it  to  the  North  American  representatives  of  this 
genus. 

The  description  of  skull  morphology  is  based  on  four  specimens;  NMNHP  Luberon  156,  NMNHP  ‘un- 
numbered’ (illustrated  by  Gaudry  1873,  pi.  6,  fig.  1,  and  Skinner  and  MacFadden  1977,  text-fig.  3a),  BMNH 
M33603,  and  BMNH  M26617  (three  of  these  are  illustrated  in  text-fig.  1). 

The  skull  is  mesocephalic  and  of  moderate  size.  The  premaxillary  and  nasal  regions  are  preserved  in  one 
specimen,  BMNH  M26617  (text-fig.  lc).  It  is  unfortunate  that  in  BMNH  M26617  the  nasal  region  is  covered 
with  matrix  and  therefore  it  is  difficult  to  determine  the  posterior  extent  of  the  premaxillary  bone  and  nasal 
notch.  However,  the  reconstructed  nasal  region  in  this  specimen  suggests  a well-retracted  nasal  notch.  In  the  four 
skulls  studied  the  buccinator  fossa  is  either  not  preserved  or  it  is  covered  with  reconstructive  material  and 
therefore  nothing  can  be  said  about  the  development  of  this  region. 


620 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  1.  Skulls  of  Hipparion  prostylum  from  the  late  Turolian  of  Mt.  Leberon,  France,  a,  NMNHP 
‘unnumbered’;  b,  NMNHP  Lub.  156;  c,  BMNH  M26617.  Shading  represents  reconstruction  or  matrix. 


M ACFADDEN:  MIOCENE  HORSE  HIPPARION 


621 


0 


2 


3 


4 


5cm 


text-fig.  2.  Deciduous  upper  cheek  teeth  (right  dP2-dP4)  of  Hipparion  prostylum,  NMNHP  Lub.  94,  from  the 
late  Turolian  of  Mt.  Leberon,  France. 

The  preorbital  facial  fossa  lies  on  the  dorsal  half  of  the  cheek  region.  Anteriorly  the  fossa  is  poorly  defined 
and  it  is  confluent  with  the  adjoining  facial  region.  Posteriorly  this  fossa  is  usually  moderately  pocketed  and  has 
a well-developed  continuous  rim.  The  fossa  lies  in  front  of  the  lacrimal  bone  (as  preserved  in  BMNH  M26617, 
text-fig.  1)  and  well  forward  of  the  orbit.  Postero-ventral  to  the  nasomaxillary  fossa  is  a moderately  developed 
malar  crest.  There  is  no  fossa  associated  with  the  malar  crest  as  is  the  case  in  some  other  horses  (e.g. 
Pliohippus).  The  teeth  are  moderately  hypsodont,  slightly  curved,  and  covered  with  cement. 

The  upper  incisors  have  cement-filled  infundibula  (cups).  The  precanine  diastema  is  smaller  than  the  post- 
canine diastema.  DP2~dP4  are  more  rectangular  in  cross-section  than  the  corresponding  P2-P4  (text-fig.  2). 
The  deciduous  premolars  are  similar  in  dental  pattern  to  the  corresponding  permanent  premolars.  In  particular, 
the  fossettes  are  moderately  plicated,  the  protocones  are  usually  rounded,  and  there  is  a tendency  for  the 
protocone  of  the  dP2  and  P2  to  become  connected  to  the  protoloph  during  relatively  early  wear  stages. 

In  the  permanent  upper  dentition  the  protocone  is  isolated  from  the  protoloph  until  late  wear  stages  (except 
in  the  P2  as  noted  above)  when  these  two  structures  frequently  connect.  The  protocone  is  characteristically 
rounded  but  infrequently  varies  to  oval  or  lenticulate  in  shape  with  anterior  and  posterior  spurs  (text-fig.  3). 
The  hypoconal  groove  is  relatively  well  developed  until  late  wear  stages.  The  enamel  plications  are  simple  to 
moderately  well  developed.  The  posterior  border  of  the  anterior  lake  (prefossette)  and  the  anterior  border  of  the 
posterior  lake  (postfossette)  show  the  most  complexity  of  plications  within  a given  tooth  or  tooth  row.  As  in 
North  American  hipparions,  the  anterior  border  of  the  prefossette  and  posterior  border  of  the  postfossette 
lack  complex  foldings.  The  plicaballin  consists  of  either  a single  or  double  loop. 


text-fig.  3.  Permanent  right  upper  cheek  teeth  (P2-M3)  of  Hipparion  prostylum,  BMNH  27590,  from  the  late 
Turolian  of  Mt.  Leberon,  France. 


0 12  3 4 5cm 


622 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  4.  Deciduous  lower  cheek  teeth  of  Hipparion  prostylum  from  the  late  Turolian  of  Mt.  Leberon,  France. 
a,  NMNHP  Lub.  14,  right  dP2-dP4;  b,  NMNHP  Lub.  26,  left  dP2-dP4. 

The  lower  incisors  have  cement-filled  infundibula.  The  precanine  diastema  is  very  small  and  the  canine  is 
nearly  appressed  to  the  I3.  The  postcanine  diastema  is  moderate  in  length,  with  the  mental  foramen  situated 
approximately  midway  between  the  C and  P2.  The  premolars  are  larger  in  cross-section  than  the  molars.  As 
exemplified  by  NMNHP  Luberon  14  and  26  (text-fig.  4),  the  lower  deciduous  teeth  are  similar  to  the 
permanent  premolars  in  dental  pattern.  In  the  anterior  region  of  dP2  and  P2  there  is  a moderately  developed 
anterior  projection  of  the  paralophid-parastylid  complex  characteristic  of  hyposodont  horses.  The  P2  through 
M3  are  generally  similar  in  dental  pattern  except  as  noted  below  (text-fig.  5).  There  is  a well-developed 
parastylid  on  the  antero-external  portion  of  the  cheek  tooth.  This  structure  is  similar  to  that  seen  in  some 
other  hipparions,  e.g.  Cormohipparion.  The  metaconids  and  metastylids  are  well  separated  and  vary  from  equal 
to  subequal  in  size.  The  entoconid  is  significantly  larger  than  the  hypoconulid.  On  the  M3  the  posterior 
portion  of  the  tooth  is  expanded  to  form  a projection  of  the  hypoconulid  or  ‘heel’.  The  protoconids  and  hypo- 
conids  are  crescentic.  In  contrast  to  e.g.  Neohipparion  eurystyle  and  Pleistocene  hipparions  from  Africa,  the 
ectoflexid  is  moderately  developed  in  the  premolars.  In  the  molars  the  deep  ectoflexid  almost  separates  the 
metaconid  and  metastylid.  The  plicaballinid  and  other  enamel  plications  are  usually  absent  or  infrequently  they 
are  poorly  developed. 


0 12  3 4 5 


cm 


text-fig.  5.  Permanent  left  lower  cheek  teeth  (P2-M2)  of  Hipparion  prostylum,  NMNHP  Lub.  40,  from  the  late 
Turolian  of  Mt.  Leberon,  France. 


M ACFADDEN:  MIOCENE  HORSE  HIPPARION 


623 


The  metapodials  of  H.  prostylum  from  Mt.  Leberon  are  of  moderate  size  relative  to  other  Eurasian 
hipparions.  Sondaar  (1974)  studied  the  metapodials  of  H.  prostylum  and  concluded  that  this  species  was 
smaller  than  the  slender  form  from  Pikermi,  H.  gracile.  As  is  the  case  in  Eurasian  hipparions  of  Turolian 
age,  H.  prostylum  usually  has  a well-developed  ectocuneiform  facet  on  the  MT  III  (Sondaar  1974,  Sondaar,  pers. 
comm.  1979). 

Discussion.  Woodburne  and  Bernor  (1980)  and  Woodburne  (pers.  comm.  1980)  suggest  that  two 
forms  of  hipparions  are  represented  at  Mt.  Leberon.  This  assertion  is  based  on  the  fact  that,  besides 
the  facial  morphotype  described  as  Hipparion  s.s.,  Pirlot  (1956)  described  one  skull  from  the  BMNH 
collection  that  had  a well-developed  preorbital  facial  fossa.  From  his  description,  one  might  be 
concerned  that  this  skull  possibly  represented  Cormohipparion.  If  that  were  true,  then  the  validity  and 
proper  assignment  of  the  species  prostylum  to  Hipparion  would  be  questionable.  Pirlot  (1956) 
unfortunately  did  not  refer  to  the  skull  in  question  by  its  catalog  number.  I have  studied  the  BMNH 
collection,  and  unless  this  skull  has  been  lost,  it  seems  almost  certain  that  based  on  Pirlot’s 
description,  he  was  referring  to  BMNH  M26617  (text-fig.  lc).  It  is  not  necessary  to  refer  this  skull 
to  another  taxon  besides  H.  prostylum  because  BMNH  M26617  appears  to  be  the  same  facial 
morphotype  as  the  other  cranial  specimens  from  Mt.  Leberon. 


Hipparion  tehonense  (Merriam  1916),  new  combination 
Text-figs.  6-8,  13,  14 


Selected  synonymy 

1907  ? Hipparion  lenticularis  (in  part),  Gidley,  pp.  915-917.  Synonymy  restricted  to  Clarendonian 
sample  from  Texas  Panhandle. 

1918  Hipparion  lenticulare  (in  part),  Osborn,  pp.  184-185,  text-figs.  147,  148;  pi.  32,  fig.  2;  pi.  33, 
figs.  5-7.  Synonymy  restricted  to  Clarendonian  sample  from  Texas  Panhandle. 

1916  Neohipparion  gratum  tehonense,  Merriam,  pp.  118-120,  text-figs.  1-7. 

1918  Hipparion  lenticulare  Osborn,  pp.  184-185,  text-figs.  147,  148;  pi.  32,  fig.  2;  pi.  33,  figs.  5-7. 
1939  Nannippus  tehonensis  Stirton,  pp.  347-352,  text-figs.  13,  24. 

1942  Nannippus  tehonensis  Drescher,  pp.  11-15,  text-fig.  3. 

1969  Nannippus  tehonensis  Webb,  pp.  130-135. 

Type  specimen  and  locality.  UCMP  21780,  right  upper  M1?,  described  by  Merriam  (1916,  p.  1 19,  fig.  1),  Chanac 
(‘Santa  Margarita’)  Formation,  south  Tejon  Hills,  California,  early  Clarendonian. 

Diagnosis.  Characters  same  as  for  other  species  of  the  genus  Hipparion  s.s.  In  particular,  the  preorbital  facial 
fossa  is  well  developed  posteriorly,  but  anteriorly  becomes  poorly  defined  (text-figs.  6,  7).  The  nasal  notch  is 
retracted  to  a position  that  lies  above  P2.  In  addition,  H.  tehonense  is  characterized  by  very  simple  enamel 
plications  and  the  anterior  region  of  the  P2  is  not  as  well  developed  as  some  other  Hipparion  s.s. 

Referred  material.  H.  tehonense  from  the  California  localities  is  represented  by  numerous  specimens  in  the 
UCMP  and  CIT  collections  (see,  e.g.,  Merriam  1916  and  Drescher  1942).  The  Texas  occurrence  of  this  species 
is  represented  by  F:AM  74400-74585  and  also  numerous  uncatalogued  F:AM  specimens  from  MacAdams 
Quarry  (locality  17),  collected  by  the  Frick  Laboratory  between  1934-1960,  Donley  County,  Texas  Panhandle 
and  also  specimens  from  other  localities  in  Donley  County,  e.g.  AMNH  10854  (see  Osborn  1918,  pi.  32,  fig.  2). 

Distribution.  Besides  the  type  locality,  H.  tehonense  is  also  known  from  the  Orinda  Formation,  early  Claren- 
donian, San  Francisco  Bay  Area,  California,  and  the  ‘Clarendon  Beds’,  Ogallala  Group,  early  Clarendonian, 
Donley  County,  Texas. 

Description.  In  most  characters,  H.  tehonense  is  similar  to  H.  prostylum.  Only  those  characters  that  show 
certain  important  similarities  and  differences  between  H.  tehonense  and  H.  prostylum  or  characters  not  repre- 
sented in  the  hypodigm  of  H.  prostylum  will  be  discussed  here. 

The  description  of  skull  morphology  of  H.  tehonensis  is  based  on  a large  sample  from  MacAdams  Quarry, 
as  exemplified  by  F:AM  74478  (text-fig.  6a),  F:AM  74537  (text-fig.  7a),  and  AMNH  10854  (‘neotype’  of 
H.  ‘ lenticulare ’,  see  Osborn  1918,  pi.  32,  fig.  2)  from  the  ‘Clarendon  Beds’  of  the  Texas  Panhandle.  The  skull 
of  H.  tehonensis  is  small  relative  to  other  species  of  Hipparion  s.s. 


624 


PALAEONTOLOGY,  VOLUME  23 


The  premaxilla  extends  postero-dorsally  to  above  the  P2-P3.  The  nasal  notch,  which  lies  above  P2,  is  well 
retracted  in  contrast  to  other  hipparions  such  as  Neohipparion  whitneyi  (see  Osborn  1918,  pi.  32,  fig.  1)  but 
certainly  less  retracted  than  e.g.  proboscideum  (see  Sondaar  1971,  pi.  III). 

The  infraorbital  canal  lies  above  P3.  As  seen  in  H.  prostylum,  the  preorbital  facial  fossa  lies  on  the  dorsal 
half  of  the  facial  region.  Anteriorly,  the  fossa  is  poorly  defined  and  it  is  confluent  with  the  adjoining  facial 
region.  Posteriorly,  this  fossa  is  usually  moderately  pocketed  and  has  a well-developed  continuous  rim.  The  fossa 
lies  well  forward  of  the  lacrimal  bone  and  orbit.  As  evidenced  by  the  MacAdams  Quarry  sample,  there  is  no 
significant  morphological  change  in  the  preorbital  facial  fossa  during  ontogeny  (compare  text-figs.  6a  and  7a). 

The  dentition  of  H.  tehonense  is  similar  in  pattern  to  other  species  of  this  genus.  The  enamel  plications  are 
very  simple  relative  to  other  hipparions.  The  protocones  are  rounded  to  oval  and  these  structures  tend  to 
become  connected  to  the  protoloph  during  later  wear  stages,  particularly  in  the  P2.  There  are  well-developed 
parastylids,  and  the  ectoflexids  are  deep  with  few,  if  any,  plicaballinids  (text-figs.  6b,  7b,  and  8). 

Discussion.  The  large  sample  from  MacAdams  Quarry  is  assigned  to  H.  tehonense  as  defined  by  the 
topotypic  material  from  the  Tejon  Hills  based  on  the  following  distinctive  characters;  (1)  small  size 
relative  to  other  Hipparion  s.s.,  (2)  extreme  simplicity  of  the  enamel  plications,  (3)  a poorly  developed 
anterior  extension  of  the  parastyle  on  P2,  and  (4)  similar  degree  of  hypsodonty. 

Because  of  its  distinctively  small  size,  the  species  H.  tehonense  from  California  has  in  the  past  been 
assigned  to  two  different  taxa  of  small  hipparions.  Merriam  (1916)  originally  named  the  topotypic 
material  from  Tejon  Hills  a subspecies  of  the  tiny  Pseudhipparion  gratum.  Subsequent  workers  have 
assigned  tehonensis  to  Nannippus,  a genus  of  dubious  monophyletic  significance.  Skinner  and 
Hibbard  (1972,  p.  117)  stated  that:  ‘The  practice  of  assigning  all  small  forms  of  Hipparion- like 


text-fig.  6.  Adult  specimen  of  Hipparion  tehonense , F:AM  74478,  from  the  Frick  MacAdams  Quarry,  early 
Clarendonian  of  the  Texas  Panhandle,  a,  left  lateral  view  of  skull;  B,  occlusal  view  of  left  upper  dentition. 
Shading  represents  reconstruction  or  matrix. 


MACFADDEN:  MIOCENE  HORSE  HIPPARION 


625 


B 


text-fig.  7.  Immature  specimen  of  Hipparion  tehonense,  F:AM  74537,  from  the  Frick  MacAdams  Quarry,  early 
Clarendonian  of  the  Texas  Panhandle,  a,  right  lateral  view  of  skull;  b,  occlusal  view  of  dP'-dP4.  Shading 
represents  reconstruction  or  matrix. 


equids  to  Nannippus  without  careful  consideration  of  other  characters  clouds  the  relationship  of 
many  of  the  dwarf  forms  and  prevents  the  recognition  of  true  Nannippus.  For  example,  Griphippus 
[=  Pseudhipparion ] gratus,  which  has  quite  different  skull,  dental,  and  postcranial  characters,  has 
often  been  assigned  to  Nannippus.' 

Although  there  are  no  skulls  preserved  for  the  Californian  sample  of  H.  tehonense , the  MacAdams 
Quarry  specimens  clearly  demonstrate  a similarity  in  facial  morphology  with  Hipparion  s.s. 
MacFadden  and  Waldrop  (1980)  described  the  facial  morphology  of  N.  phlegon  from  Mt.  Blanco  in 
the  Texas  Panhandle,  which  is  the  genotypic  locality  and  therefore  central  to  the  concept  of  that 
genus.  N.  phlegon  has  a smooth  preorbital  cheek  region  with  no  facial  fossa.  Therefore,  there  is  no 
doubt  that  the  small  hipparion  species  tehonense  is  best  referred  to  the  genus  Hipparion  s.s. 


0 1 


4 


5 cm 


text-fig.  8.  Right  lower  cheek  teeth  (P2-M3)  of  Hipparion  tehonense,  F:AM  105440,  from  the  Frick 
MacAdams  Quarry,  early  Clarendonian  of  the  Texas  Panhandle. 


626 


PALAEONTOLOGY,  VOLUME  23 


lenticularis,  as  it  is  used  for  Clarendonian  hipparions  from  Donley  County,  Texas,  is 
synonymized  here  with  H.  tehonense.  The  species  lH.'  lenticularis  has  been  inconsistently  used  in  the 
literature  and  it  is  appropriate  to  comment  on  its  nomenclature  here.  In  1893  Cope  assigned  the 
species  lenticularis  to  Protohippus  based  on  material  of  late  Hemphillian  age  from  Mulberry  Canyon, 
near  Goodnight,  in  the  Texas  Panhandle  (see  Schultz  1977).  Gidley  (1907)  referred  material  from 
the  Clarendon  beds  of  Donley  County  in  the  Texas  Panhandle  to  H.  lenticularis.  Osborn  (1918) 
designated  a well-preserved  skull  (also  described  previously  by  Gidley  1907),  AMNH  10584,  as  the 
neotype  of  H.  lenticularis.  This  judgement  was  apparently  made  by  Osborn  because  the  early  workers 
thought  that  the  Clarendon  and  Goodnight  beds  were  correlative  and  the  topotypic  material  from 
Mulberry  Canyon  was  not  abundant.  Despite  these  previous  taxonomic  decisions,  it  remains  to  be 
demonstrated  that  ‘Hf  lenticularis  from  Donley  County  is  conspecific  with  the  material  from 
Mulberry  Canyon  and  numerous  other  late  Hemphillian  localities,  e.g.  Coffee  Ranch  (Matthew  and 
Stirton  (1930).  It  is  unfortunate  that  no  skulls  are  known  of  late  Hemphillian  lenticularis.  The 
Clarendonian  H.  tehonense  and  Hemphillian  lenticularis  are  remarkably  similar  in  dental 
pattern,  however,  the  younger  species  is  noticeably  more  hypsodont.  In  this  report  the  species 
lenticularis  is  restricted  to  the  late  Hemphillian  forms.  Based  on  dental  and  temporal  similarities, 
the  Clarendonian  lenticularis  as  used  by  workers  such  as  Gidley  and  Osborn  is  synonymized  with 
H.  tehonense. 


Hipparion  forcei  Richey  1948 
Text-figs.  9,  13,  14 

Selected  synonymy 

1919  Hipparion  mohavense  Merriam,  pp.  549-553,  text-figs.  163-170. 

1948  Hipparion  forcei  Richey,  pp.  9-25,  text-figs.  4-12,  pi.  2,  figs,  a-c;  pi.  3,  figs.  a-d. 

1969  Nannippus  forcei  Webb,  pp.  130-135. 

Type  specimen  and  locality.  UCMP  33051,  P3,  from  Green  Valley  Formation,  Black  Hawk  Ranch  Quarry, 
Mount  Diablo  area,  California,  late  Clarendonian. 

Diagnosis.  Characters  same  as  for  other  species  of  the  genus  Hipparion  s.s.,  in  particular,  configuration  of  the 
preorbital  facial  fossa  and  nasal  region  listed  above  for  H.  prostylum  and  H.  tehonense.  Specific  characters  for 
H.  forcei  include  an  apparently  higher  frequency  of  connection  of  the  protocone  to  the  protoloph  in  P2 
(Richey  1948).  Also  the  protocone-protoloph  connection  is  very  well  developed  with  less  of  a constriction 
between  these  parts  than  is  seen  in  many  other  hipparions.  The  protocone  is  smaller  in  H.  forcei  than  in 
H.  tehonense  relative  to  the  occlusal  area  of  the  tooth.  H.  forcei  has  higher  crowned  cheek  teeth  with  larger 
occlusal  cross-sectional  areas  than  in  H.  tehonense  (Webb  1969). 

Referred  material.  Numerous  UCMP  specimens  from  the  Black  Hawk  Ranch  Local  Fauna,  Green  Valley 
Formation,  San  Francisco  Bay  region,  California,  and  the  Dove  Springs  Fauna,  Ricardo  Formation,  Mohave 
Desert,  California  (see  Richey  1948). 

Distribution.  Besides  the  type  locality,  H.  forcei  is  probably  represented  in  the  Ricardo  Formation  (Dove 
Springs  Fauna),  Mohave  Desert,  California.  These  localities  are  late  Clarendonian  in  age  (Tedford  et  al.,  in 
press). 

Description.  In  most  characters  the  material  of  H.  forcei  is  similar  to  other  species  of  this  genus,  including 
H.  prostylum  and  H.  tehonense.  The  cranial  morphology  of  H.  forcei  is  known  from  one  crushed  but  relatively 
complete  skull  from  Black  Hawk  Ranch  (text-fig.  9a),  UCMP  34511,  originally  described  in  detail  by  Richey 
(1948).  The  important  characters  that  are  similar  among  these  species  include  a relatively  well-developed  nasal 
notch  that  is  retracted  to  a position  that  lies  over  P2.  The  infraorbital  canal  lies  above  P3.  As  is  diagnostic  of 
Hipparion  s.s.,  the  preorbital  facial  fossa  is  poorly  defined  anteriorly  but  posteriorly  it  is  characterized  by  a well- 
developed  continuous  rim  that  is  pocketed.  The  fossa  lies  well  forward  of  the  lacrimal  bone  and  orbit. 

The  most  complete  dentition  of  H.  forcei  is  known  from  the  skull,  UCMP  34511  (text-fig.  9b).  However,  the 
dental  pattern  in  this  specimen  is  not  characteristic  because  it  represents  an  old  individual  in  late  wear  stage. 
There  are  numerous  isolated  teeth  known  from  the  type  locality  and  Richey  (1948)  described  them  in  detail. 


MACFADDEN:  MIOCENE  HORSE  HIPPARION 


627 


The  following  characters  are  diagnostic  of  H.  forcer,  relatively  simple  enamel  plications,  small  protocone,  high 
frequency  of  protocone-protoloph  connection  in  the  P2,  and  lowers  with  deep  ectoflexids  but  without 
plicaballinids.  Richey  (1948,  p.  15)  stated  that:  ‘Another  character  that  distinguishes  H.forcei  from  many  other 
species  is  the  frequency  of  connection  of  the  protocone  with  the  protoconule  [protoloph],  Many  hipparions 
have  a connected  protocone  in  the  P2.  This  is  particularly  true  of  H.  forcei.  In  fact,  in  none  of  the  specimens  thus 
far  studied  is  the  protocone  separate.’ 

Richey  (1948)  studied  the  limbs  of  H.  forcei  and  concluded  that  they  were  of  moderate  size  in  contrast  to 
smaller  forms  such  as  Nannippus  and  larger,  more  robust,  forms  such  as  ‘ H .’  (=  Cormohipparion)  theobaldi 
from  the  Siwaliks  and  H.  gracile  from  Pikermi. 


text-fig.  9.  Hipparion  forcei,  UF  22656  (cast  of  UCMP  3451 1)  from  the  late  Clarendonian  Black  Hawk  Ranch 
Local  Fauna,  California.  A,  left  lateral  view  of  skull;  b,  occlusal  view  of  right  upper  cheek  teeth. 


Discussion.  H.forcei  and  H.  tehonense  are  very  similar  in  many  characters.  Webb  (1969)  has  suggested 
an  ancestral-descendent  relationship  between  these  two  species.  The  samples  from  Tejon  Hills- 
Chanac  Formation-Black  Hawk  Ranch  appear  to  approximate  a morphocline  in  characters  such  as 
hypsodonty.  However,  in  other  characters  such  as  the  high  frequency  of  protocone-protoloph  con- 
nection, H.forcei  seems  more  primitive  than  H.  tehonense.  The  relative  primitiveness  of  certain  dental 
characters  in  H.forcei  would,  as  Richey  (1948)  suggested,  imply  independent  evolution  in  parallel  of 
H.  forcei  and  H.  tehonense  from  a common  ancestor  rather  than  a single  ancestral-descendent 
sequence  as  suggested  by  Webb  (1969).  It  is  not  within  the  scope  of  this  paper  to  resolve  the  phylo- 
genetic relationships  of  the  species  H.  tehonense  and  H.forcei.  This  short  note  is  included  as  an  intro- 
duction to  the  next  section  below,  i.e.  the  provisional  assignment  of  forms  from  the  mid-continent 
of  North  America  to  H.  cf.  tehonense  or  forcei. 


628 


PALAEONTOLOGY,  VOLUME  23 


Hipparion  cf.  tehonense  or  forcei 
Text-figs.  10-14 

Referred  material.  Numerous  specimens  in  the  F:AM  collection  including  well-preserved  skulls,  e.g.  F:AM 
107664,  from  Trail  Side  Kat  Quarry  Channel,  Cherry  County,  Nebraska,  late  Clarendonian;  F:AM  107663, 
Rosebud  Agency  Quarry,  Todd  County,  South  Dakota,  late  Clarendonian;  F:AM  71887,  Olcott  Quarry, 
Hipparion  Channel,  Olcott  Hill,  Sioux  County,  Nebraska,  late  Clarendonian. 

Distribution.  Snake  Creek  and  Ash  Hollow  Formations,  Ogallala  Group,  north-central  Nebraska  and  adjacent 
South  Dakota,  and  north-western  Nebraska,  late  Clarendonian  (see  Skinner  et  al.  1977;  Tedford  et  al.  in  press). 

Description.  Well-preserved  cranial  material  from  the  northern  Great  Plains  localities  are  referred  to  Hipparion 
s.s.  based  on  the  configuration  of  certain  skull  characters,  particularly  the  preorbital  facial  fossa. 

There  appears  to  be  a significant  size  difference  among  the  individuals  of  H.  cf.  tehonense  or  forcei.  In 
F:AM  107664  (text-fig.  10)  and  F:AM  107663  (text-fig.  11)  the  premaxillary  extends  posteriorly  to  a position 
that  lies  over  the  P2.  There  is  some  variation  in  the  posterior  extent  of  the  nasal  notch.  In  F:AM  107664 
and  F:AM  71887  (text-fig.  12)  the  nasal  notch  extends  to  a position  that  lies  over  the  buccinator  fossa,  which 
is  slightly  less  retracted  than  in  other  skulls  of  Hipparion  s.s.  described  here.  Although  the  nasal  bones  are  not 
preserved  in  F:AM  107664,  the  nasal  notch  appears  retracted  to  a position  over  P2  similar  to  that  seen  in  other 
skulls  of  Hipparion  s.s.  In  the  skulls  illustrated  in  text-figs.  10-12  the  infraorbital  foramen  lies  above  the  P3 
just  ventral  to  the  antero-ventral  margin  of  the  preorbital  facial  fossa.  This  fossa  is  poorly  defined  anteriorly  but 
posteriorly  it  consists  of  a well-defined  continuous  rim.  Posteriorly  there  also  is  a moderately  well-developed 
pocket.  This  fossa  lies  well  forward  of  the  lacrimal  bone  and  orbit.  There  is  a moderately  developed  malar  crest. 

The  dentitions  are  similar  to  other  species  of  Hipparion  s.s.  In  particular,  the  enamel  plications  are  relatively 
simple.  The  protocone  is  oval  and  relatively  large.  In  the  P2  of  F:AM  107663  and  F:AM  107664  the  protocone 
is  strongly  connected  to  the  protoloph. 


text-fig.  10.  Hipparion  cf.  tehonense  or  forcei,  F:AM  107664,  from  the  late  Clarendonian  Trail  Side  Kat  Quarry 
Channel,  Nebraska,  a,  left  lateral  view  of  skull;  B,  occlusal  view  of  left  upper  cheek  teeth. 


M ACFADDEN:  MIOCENE  HORSE  HIPPARION 


629 


B 


text-fig.  11.  Hipparion  cf.  tehonense  or  forcei,  F:AM  107663,  from  the  late  Clarendonian  Rosebud  Agency 
Quarry,  South  Dakota,  a,  left  lateral  view  of  skull;  b,  occlusal  view  of  left  upper  cheek  teeth. 


text-fig.  12.  Hipparion  cf.  tehonense  or  forcei,  F:AM  71887,  from  the  late  Clarendonian  Olcott  Quarry, 
Hipparion  Channel,  Olcott  Hill,  Nebraska,  a,  left  lateral  view  of  skull;  b,  occlusal  view  of  left  upper  cheek  teeth. 


630 


PALAEONTOLOGY,  VOLUME  23 


Discussion.  The  configuration  of  the  skull,  particularly  in  the  development  of  the  preorbital  facial 
fossa,  justifies  the  allocation  of  the  material  from  these  mid-continental  sites  to  Hipparion  s.s.  How- 
ever, the  specific  allocation  is,  at  this  point,  somewhat  uncertain.  It  is  not  implied  that  the  sample 
from  these  three  localities  represents  one  discrete  species.  For  example,  the  smaller  size  of  F:AM 
107664  and  F:AM  107663  possibly  indicates  an  affinity  with  H.  tehonense,  whereas  the  larger  size  of 
F:AM  71887  possibly  indicates  an  affinity  with  H.forcei  (following  Webb  1969).  On  the  other  hand, 
the  very  strong  connection  of  the  protocone  and  protoloph  in  both  F:AM  107663  and  F:AM 
107664  indicates  an  affinity  with  H.  forcei  (following  Richey  1948).  The  resolution  of  this  species- 
level  problem  would  require  further  study  beyond  the  scope  of  the  present  paper.  The  important 
point  is  that  this  mid-continental  sample  is  referred  to  Hipparion  s.s.  Therefore,  this  genus  was 
relatively  widespread  in  North  America  during  the  Clarendonian. 


BIOSTRATIGRAPHY  AND  PALAEOBIOGEOGR APH Y 

The  temporal  and  geographic  distribution  of  Hipparion  s.s.  from  North  America  and  Mt.  Leberon  is 
summarized  in  text-figs.  13  and  14.  In  these  text-figs,  several  other  Eurasian  localities  of  Hipparion 
s.s.  have  been  added,  based  on  recent  studies  of  cranial  morphology  (MacFadden  and  Bakr  1979, 
Woodburne  and  Bernor  1980).  Undoubtedly  Hipparion  s.s.  (as  recognized  by  cranial  morphology) 
occurs  at  other  Holarctic  localities  and  possibly  also  in  Africa.  For  this  discussion  only  the  selected 
localities  shown  in  text- figs.  13  and  14  will  be  presented. 

The  radiometric  time  scale  and  European  Stages  in  text-fig.  14  were  taken  from  several  works, 
including  Berggren  and  Van  Couvering  (1974),  Aguirre  (1975),  Fahlbusch  (1976),  and  Van 
Couvering  and  Berggren  (1977).  The  age  of  the  ‘Hipparion  Datum  Plane’  is  shown  to  range  from 
about  12  0 mybp  to  about  10-8  mybp.  This  range  is  a result  of  alternative  interpretations  of  radio- 
metric  dating  of  critical  Old  World  sites,  particularly  Howenegg  (e.g.  Berggren  and  Van  Couvering 
1974;  Van  Couvering  and  Berggren  1977;  Becker-Platen  et  at.  1977;  Barndt  et  al.  1978,  also  see 
discussion  in  MacFadden  and  Bakr  1979).  The  boundary  between  the  Astarcian  ( sensu  Fahlbusch 
1976)  and  Vallesian,  which  is  usually  taken  as  the  first  appearance  of  ‘Hipparion’,  is  dashed  in 
text-fig.  14  in  order  to  indicate  the  uncertainty  involved  in  the  calibration  of  this  event.  The  calibra- 
tion and  nomenclature  of  the  North  American  localities  is  taken  from  Tedford  et  al.  (in  press). 

Sondaar  (1974)  and  Sen  et  al.  (1978)  consider  the  Mt.  Leberon  locality,  and,  therefore,  the  type 
material  of  Hipparion  s.s.  to  be  of  late  Turolian  age.  Woodburne  and  Bernor  (1980)  studied  facial 
morphotypes  from  selected  Old  World  localities.  Their  ‘Group  3’  consists  of  a morphologically 
distinct  group  including  H.  prostylum  from  Mt.  Leberon  and  forms  from  several  localities  in  Greece 
and  Iran.  This  facial  morphotype  agrees  with  the  concept  of  the  genus  presented  in  this  report  for 
Hipparion  s.s.  The  Hipparion  s.s.  from  Saloniki,  Greece  (Group  3 of  Woodburne  and  Bernor  1980, 
here  referred  to  as  H.  ‘ prostylum ’),  and  Pikermi,  Greece  (also  Group  3,  here  referred  to  as  H.  ‘ gracile ’), 
are  considered  to  be  of  medial  Turolian  age  and  slightly  older  than  Mt.  Leberon  (Berggren  and  Van 
Couvering,  1974;  Sen  et  al.  1978). 

For  a long  time  it  was  thought  that  at  Samos,  Quarries  1-4  were  older  than  Quarry  5,  and  that  this 
succession  spanned  medial  to  late  Turolian  time  (e.g.  Berggren  and  Van  Couvering  1974;  Aguirre 
1975;  Sen  et  al.  1978).  Recent  field  work  at  Samos  (Solunias,  pers.  comm.  1977),  suggests  that  all  the 
quarries  are  approximately  contemporaneous.  Therefore,  depending  upon  the  stratigraphic  inter- 
pretation, the  Hipparion  s.s.  (Group  3,  here  referred  to  as  H.  ‘ dietrichC ) from  Samos  either  spans 
medial  to  late  Turolian  time  or  is  restricted  to  the  late  Turolian.  Woodburne  and  Bernor  (1980)  state 
that  Hipparion  Group  3 (here  referred  to  H.  sp.)  is  found  in  the  middle  and  upper  parts  of  the 
Maragheh,  Iran,  sequence.  Based  on  this  range,  the  Maragheh  Hipparion  s.s.  spans  medial  to  late 
Turolian  time. 

The  Siwalik  hipparions  of  Pakistan  and  adjacent  India  have  been  the  subject  of  numerous  publica- 
tions because  of  their  association  with  a very  rich  Neogene  sequence  including  the  oldest-known 
hominoid  fossils  (Pilbeam  et  al.  1977).  Hussain  (1971)  presented  the  most  recent  revision  of  Siwalik 
hipparions.  MacFadden  and  Bakr  (1979)  recognize  two,  or  perhaps  three,  supraspecific  taxa  of 


MACFADDEN:  MIOCENE  HORSE  HIPPARION 


631 


MacFadden  and  Bakr  (1979)  and  Woodburne  and  Bernor  (1980).  North  American  localities  (7-13)  are  taken  from  this  report.  The  locality 
numbers  in  this  text-fig.  correspond  to  the  numbers  of  the  columns  in  text-fig.  14. 


v 

SELECTED  HOLARCTIC  LOCALITIES  OF  HIPPARION  S-  S. 

EURASIA 

NORTH  AMERICA 

*S» 

ass 

tea 

a 

cSS± 

* 

F 

.ffiSn 

13 

/ats 

i 

c 

E ^ 
> 

. 

rn 

-12 

A« 

t 

\ 

1 

JL. 

- 

— 

2 

JL 

i 

i 

s 

1 

\ 

632 


PALAEONTOLOGY,  VOLUME  23 


z <->?*< 

— — i— i — <Z  u-^<aiuZQOZ-<Z  <°^ 

CD 

SELECTED  HOLARCTIC  LOCALITIES  OF  HIPPARION  S.  S. 

NORTH  AMERICA 

13 

Southern 
S.  Dakota 

H. 

cf. 

tehonense 

or 

forcei 

12 

North- 

Central 

Nebraska 

H. 

cf. 

tehonense 

or 

forcei 

11 

Sioux 

County, 

Nebraska 

H. 

cf. 

tehonense 

or 

forcei 

10 

Clarendon 

Beds, 

Texas 

H. 

tehonense 

9 

S.Tejon 

Hills, 

California 

H. 

tehonense 

8 

Black  Hawk 
Ranch, 
California 

H. 

forcei 

7 

Orinda, 

California 

H. 

tehonense 

EURASIA 

6 

Siwaliks, 

Pakistan 

4 

i 

H. 

cf. 

antilo- 

pinum 

1 

1 

? 

i 

l 

t 

5 

Maragheh, 
1 ran 

H. 

sp- 

NE 

4 

Samos, 

Greece 

H. 

"dietrichi" 

7 

\ 

SIAN 

UUM  PLA 

3 

Pikermi, 

Greece 

H. 

"gracile" 

EURA 
’ARION  D/ 

2 

Saloniki, 

Greece 

'E  a- 

x 

1 

Mt. 

Leberon, 

France 

H. 

prostylum 

T 

M 

E 

M. 

Y. 

B. 

P. 

O IV  OO  O'  2 — — 

till  II 

. </>i—  <Ouj</> 

lU^OC 

ttIUlOU  •-  Z>£*0— ' — <Z 

uja-OUI 

M 

O 

c 

E 

N 

E 

text-fig.  14.  Temporal  and  geographic  distribution  of  the  Holarctic  Hipparion  s.s.  localities  shown  in  text-fig.  13.  Epoch,  Stage,  North 
American  Land  Mammal  ‘Ages’  (NALMA),  and  time  (mybp)  calibrations  are  taken  from  numerous  references  cited  in  the  text.  Dashed  zone 
between  Astarcian  and  Vallesian  European  Stages  indicates  the  uncertainty  involved  in  the  calibration  of  the  Hipparion  Datum  Plane.  The 
arrows  in  columns  4 and  6 indicate  questionable  ranges  depending  upon  stratigraphic  interpretations  (for  Samos)  and  lack  of  well-preserved 

cranial  material  (for  the  Siwaliks).  See  discussion  in  text. 


M ACFADDEN:  MIOCENE  HORSE  HIPPARION 


633 


Siwalik  hipparions.  Some  specimens  of  their  ‘small  hipparion  complex’  (here  termed  H.  cf.  anti- 
lopinum)  are  tentatively  referred  to  Hipparion  s.s.  Based  on  teeth,  hipparions  are  known  to  range  in 
the  Siwaliks  from  the  early  Vallesian  (roughly  10-5  mybp  following  Barndt  et  al.  1978)  to  the  early 
Villafranchian  (roughly  3 0 mybp  following  Keller  et  al.  1977).  It  is  impossible  at  present  to  deter- 
mine the  exact  range  of  H.  cf.  antilopinum  because  the  relevant  cranial  material  either  has  poor  strati- 
graphic data  (particularly  from  the  early  collections)  or  is  limited  to  the  upper  Dhok  Pathan 
Formation,  which  is  probably  late  Turolian  in  age. 

The  stratigraphic  range  in  North  America  of  Hipparion  s.s.  as  recognized  by  cranial  morphology 
is  from  early  to  late  Clarendonian.  The  individual  ages  of  each  locality  are  represented  in  text- 
figs.  13  and  14. 

There  are  several  important  conclusions  based  on  the  present  study  of  Hipparion  s.s.  First,  in 
contrast  to  the  hypotheses  of  earlier  workers,  Hipparion  s.s.  was  widespread  in  North  America  as  well 
as  Eurasia  during  the  Miocene.  The  stratigraphic  distribution  (text-fig.  14)  of  the  several  species  of 
Hipparion  s.s.  demonstrates  a generic-level  continuity  throughout  Holarctica  during  the  medial  to 
late  Turolian  and  early  to  late  Clarendonian.  The  slightly  older  occurrences  of  H.  tehonense  in  North 
America  may  be  significant  in  a phylogenetic  context  depending  upon  the  accuracy  of  the  inter- 
continental correlations.  If  H.  tehonense  is  older  than  the  other  Hipparion  s.s.,  at  present  there  is  no 
implication  of  primitiveness  or  ancestry  for  this  species.  The  interspecific  relationships  of  the  species 
assigned  to  Hipparion  s.s.  require  a detailed  analysis  beyond  the  scope  of  this  paper. 

Another  interesting  conclusion,  based  on  the  limited  number  of  localities  and  relevant  cranial 
material  discussed  here,  is  that  Eurasian  Hipparion  s.s.  appears  to  be  common  in  Turolian  age 
localities  but  is  not  recognized  from  the  Vallesian.  Therefore,  it  appears  that  Hipparion  s.s.  was  not 
involved  in  the  Eurasian  ‘Hipparion  Datum  Plane’  that  defines  the  base  of  the  Vallesian.  MacFadden 
and  Skinner  (1977)  and  Skinner  and  MacFadden  (1977)  suggested  that  the  presence  of  hipparions  in 
the  Old  World  could  have  been  a result  of  more  than  one  dispersal  event  rather  than  only  one  event 
as  has  been  suggested  by  some  other  workers  (e.g.  Forsten  1968;  Hussain  1971). 

As  was  stated  in  the  Introduction,  the  species-level  taxonomy  of  Holarctic  Hipparion  s.s.  needs  to 
be  revised  in  light  of  cranial  morphology.  Based  on  numerous  cranial  characters  it  appears  that 
Hipparion  s.s.  is  composed  of  a monophyletic  group  of  several  species.  It  is  important  to  determine 
the  ancestral  stock  from  which  the  Hipparion  s.s.  species  were  descended.  It  is  clear  that  the  closest 
relative  of  Hipparion  s.s.  is  within  the  merychippine  horses  (e.g.  Matthew  1924;  Colbert  1935;  Stirton 
1940;  Forsten  1968;  Skinner  and  MacFadden  1977).  A study  is  needed  that  demonstrates  the 
relatedness  of  several  hipparion  groups  that  apparently  arose  independently  from  the  horizontal 
merychippine  complex.  A striking  consequence  of  recent  studies  of  hipparion  cranial  morphology  is 
that  these  three-toed  horses  are  certainly  polyphyletic  and  arose  from  more  than  one  merychippine 
lineage.  In  short,  several  distinct  supraspecific  groups  of  hipparions  originated  independently  and 
evolved  in  parallel  during  the  Neogene  in  the  Old  and  New  Worlds. 


Acknowledgements.  Special  thanks  are  extended  to  Drs.  Morris  F.  Skinner  and  Richard  H.  Tedford  of  the 
American  Museum  of  Natural  History  for  their  expertise  and  guidance  during  my  studies  of  equid  palaeontology. 
I also  thank  Dr.  Tedford  for  allowing  some  illustrations  drawn  at  the  American  Museum  of  Natural  History 
to  be  used  in  this  study.  Anthony  J.  Sutcliffe,  Jeremy  Hooker,  and  Andrew  Currant  greatly  aided  in  my  studies 
of  the  hipparions  at  the  British  Museum  (Natural  History).  Many  other  colleagues  have  contributed  to  the 
present  study,  including  Vera  Eisenmann,  Paul  Y.  Sondaar,  S.  David  Webb,  Ronald  G.  Wolff,  and  Michael 
O.  Woodburne.  Nancy  Halliday  and  Lauren  Keswick  skilfully  prepared  many  of  the  illustrations.  The 
manuscript  was  typed  by  Mrs.  S.  Sidaway.  This  report  is  University  of  Florida  Contribution  to  Vertebrate 
Paleontology  number  179. 


634 


PALAEONTOLOGY,  VOLUME  23 


REFERENCES 

aguirre,  e.  1975.  Division  estratigrafica  del  Neogeno  continental.  Estudios  Geol.  31,  587-595. 
barndt,  j.  et  al.  1978.  The  magnetic  polarity  stratigraphy  and  age  of  the  Siwalik  Group  near  Dhok  Pathan 
village,  Potwar  Plateau,  Pakistan.  Earth  Planet.  Sci.  Letters,  41,  355-364. 
becker-platen,  ].  d.,  benda,  l.  and  steffans,  p.  1977.  Litho-  und  biostratigraphische  Deutung  radiometrischer 
Altersbestimmungen  aus  dem  Jungtertiar  de  Tiirkei  (Kanozoikum  und  Braunkohlen  der  Tiirkei,  18).  Geol. 
Jb.  B25,  139-167. 

berggren,  w.  a.  and  van  couvering,  J.  a.  1974.  The  late  Neogene:  Biostratigraphy,  geochronology,  and 
paleoclimatology  of  the  last  15  million  years  in  marine  and  continental  sequences.  Palaeogeogr., 
Palaeoclimat.,  Palaeoecol.  16,  1-216. 

de  christol,  j.  1832.  [No  title:  Description  of  Hipparion .]  Sci.  lnd.  Ann.  Midi,  France,  1,  180-181. 

Colbert,  e.  h.  1935.  Siwalik  mammals  in  the  American  Museum  of  Natural  History.  Trans.  Amer.  Phil.  Soc. 
N.s.  26,  1-401. 

cope,  E.  D.  1893.  A preliminary  report  on  the  vertebrate  paleontology  of  the  Llano  Estacadao.  Texas  Geol. 
Survey,  4th  Ann.  Report.  Pp.  1-137. 

drescher,  a.  b.  1942.  Later  Tertiary  Equidae  from  the  Tejon  Hills,  California.  Contrib.  Paleont.,  Carnegie 
Inst.,  Washington,  530,  1-23. 

fahlbusch,  v.  1976.  Report  on  the  International  Symposium  on  mammalian  stratigraphy  of  the  European 
Tertiary.  Newsl.  Strat.  5,  160-167. 

forsten,  a.-m.  1968.  Revision  of  Palearctic  Hipparion.  Acta  Zool.  Fennica,  119,  1-134. 
gaudry,  a.  1873.  Animaux  fossiles  du  Mont  Leberon  ( Vaucluse) : Etude  sur  les  vertebres.  F.  Savy  (ed.),  Libr.  Soc. 
Geol.  France.  Pp.  1-112. 

gervais,  p.  1949.  Note  sur  la  multiplicity  des  especes  d’hipparion  (genre  de  chevaux  a trois  droits)  qui  sont 
enfouis  a Cucuron  {Vaucluse).  Compt.  Rend.  Acad.  Sci.,  Paris,  29,  284-286. 
gidley,  j.  w.  1903.  A new  three-toed  horse.  Bull.  Amer.  Mus.  Nat.  Hist.  19,  465-476. 

— 1907.  Revision  of  the  Miocene  and  Pliocene  Equidae  of  North  America.  Ibid.  23,  865-934. 
gray,  j.  e.  1821.  On  the  natural  arrangement  of  vertebrose  animals.  London  Med.  Repository  Rev.  15,  296-310. 
Gromova,  v.  1955.  Le  genre  Hipparion.  Inst.  Paleont.  Acad.  Sci.  URSS  36.  [Original  in  Russian.]  French  transla- 
tion by  St.  Aubin,  Bur.  Rech.  Min.  Geol.  Ann.  C.E.D.P.  12,  1-473. 
hussain,  s.  t.  1971.  Revision  of  Hipparion  (Equidae,  Mammalia)  from  the  Siwalik  Hills  of  Pakistan  and  India. 
Verlag.  Bayer.  Akad.  Wiss.  N.s.  147,  1-68. 

keller,  H.  m.  et  al.  1977.  Magnetic  polarity  stratigraphy  of  the  Upper  Siwalik  deposits,  Pabbi  Hills,  Pakistan. 
Earth  Planet.  Sci.  Letters , 36,  187-201. 

linnaeus,  c.  1758.  Sy sterna  Naturae  per  Regna  Tria  Naturae,  Secundum  Classes,  Or  dines.  Genera,  Species  cum 
Characteribus  Differentiis  Synonymis  Locis.  Editia  decima,  reformata.  Laurentii,  Stockholm,  Salvi,  1,  1-824. 
macfadden,  b.  J.  and  bakr,  A.  1979.  The  horse  Cormohipparion  theobaldi  from  the  Neogene  of  Pakistan,  with 
comments  on  Siwalik  hipparions.  Palaeontology,  22,  439-447. 

— and  skinner,  m.  f.  1977.  Earliest  known  Hipparion  from  Holarctica.  Nature,  London,  265,  532-533. 

— and  waldrop,  j.  s.  1980.  Nannippus  phlegon  (Mammalia,  Equidae)  from  the  Pliocene  (Blancan)  of  Florida. 
Bull  Fla.  State  Mus.,  Biol.  Sci.  25,  1-37. 

matthew,  w.  d.  1924.  Third  contribution  to  the  Snake  Creek  fauna.  Bull.  Amer.  Mus.  Nat.  Hist.  50,  59-210. 

and  stirton,  r.  a.  1930.  Equidae  from  the  Pliocene  of  Texas.  Univ.  California  Pub.  Bull.  Dept.  Geol.  Sci. 

19,  349-396. 

merriam,  J.  c.  1916.  Mammalian  remains  from  the  Chanac  Formation  of  the  Tejon  Hills,  California.  Ibid.  10, 
111-127. 

— 1919.  Tertiary  mammalian  faunas  of  the  Mohave  Desert.  Ibid.  11,  437-585. 
osborn,  h.  f.  1918.  Equidae  of  the  Oligocene,  Miocene,  and  Pliocene  of  North  America.  Iconographic  type 
revision.  Mem.  Amer.  Mus.  Nat.  Hist.  N.s.  2,  1-326. 

owen,  R.  1948.  Description  of  teeth  and  portions  of  jaws  of  two  extinct  anthracotheroid  quadrupeds 
{Hyopotamus  vectianus  and  H.  bovinus)  discovered  by  the  Marchioness  of  Hastings  in  the  Eocene  deposits  on 
the  N.W.  coast  of  the  Isle  of  Wight,  with  an  attempt  to  develop  Cuvier’s  idea  of  the  classification  of 
pachyderms  by  the  number  of  toes.  Quart.  Jour.  Geol.  Soc.  5,  380-383. 
pilbeam,  d.  r.  et  al.  1977.  Geology  and  palaeontology  of  Neogene  strata  of  Pakistan.  Nature,  London,  270, 
684-689. 

pirlot,  p.  r.  1956.  Les  formes  Europeennes  du  genre  Hipparion.  Mem.  y Comun.  Inst.  Geol.  Barcelona,  14, 
1-122. 


MACFADDEN:  MIOCENE  HORSE  HIPPARION 


635 


richey,  k.  A.  1948.  Lower  Pliocene  horses  from  Black  Hawk  Ranch,  Mount  Diablo,  California.  Univ.  California 
Pub.  Bull.  Dept.  Geol.  Sci.  28,  1-44. 

schultz,  G.  r.  (ed.)  1977.  Guidebook:  Field  conference  on  late  Cenozoic  biostratigraphy  of  the  Texas 
Panhandle  and  adjacent  Oklahoma,  4-6  August  1977.  Kilgore  Res.  Center,  Dept.  Geol.  and  Anthropol.,  West 
Texas  State  Univ.,  Spec.  Pub.  1,  1-160. 

sen,  s.,  sondaar,  p.  Y.  and  staesche,  u.  1978.  The  biostratigraphical  application  of  the  genus  Hipparion  with 
special  references  to  the  Turkish  representatives.  Proc.  Kon.  Ned.  Akad.  Wet.  B81,  370-385. 

skinner,  m.  f.  and  hibbard,  c.  w.  1972.  Early  Pleistocene  pre-glacial  and  glacial  rocks  and  faunas  of  north- 
central  Nebraska.  Bull.  Amer.  Mus.  Nat.  Hist.  148,  1-148. 

— and  macfadden,  B.  J.  1977.  Cormohipparion  N.  Gen.  (Mammalia,  Equidae)  from  the  North  American 
Miocene  (Barstovian-Clarendonian).  J.  Paleont.  51,  912-926. 

skinner,  s.  M.  and  gooris,  R.  j.  1977.  Stratigraphy  and  biostratigraphy  of  late  Cenozoic  deposits  in 

central  Sioux  County,  western  Nebraska.  Bull.  Amer.  Mus.  Nat.  Hist.  158,  263-370. 

— and  taylor,  B.  E.  1967.  A revision  of  the  geology  and  paleontology  of  the  Bijou  Hills,  South  Dakota. 
Amer.  Mus.  Nat.  Hist.  Novitates,  2300,  1-53. 

stirton,  r.  a.  1939.  Cenozoic  mammal  remains  from  the  San  Francisco  Bay  region.  Univ.  California  Pub.  Bull. 
Dept.  Geol.  Sci.  24,  339-410. 

1940.  Phylogeny  of  North  American  Equidae.  Ibid.  25,  165-198. 

— 1941.  Development  of  characters  in  horse  teeth  and  the  dental  nomenclature.  Jour.  Mammal,  22, 434-446. 

sondaar,  p.  y.  1971.  The  Samos  Hipparion.  Proc.  Kon.  Ned.  Akad.  Wet.  B74,  417-441. 

1974.  The  Hipparion  of  the  Rhone  Valley.  Geobios,  7,  289-306. 

tedford,  r.  h.  et  al.  (in  press).  Faunal  succession  and  biochronology  of  the  Arikareean  through  Hemphillian 
interval  (late  Oligocene  through  late  Miocene  Epochs),  North  America.  Univ.  California  Press,  Berkeley. 

van  couvering,  j.  a.  and  berggren,  w.  a.  1977.  Biostratigraphical  basis  of  the  Neogene  time  scale.  In 
kauffman,  E.  G.  and  hazel,  j.  e.  (eds.),  Concepts  and  Methods  of  Biostratigraphy.  Dowden,  Hutchinson,  and 
Ross,  Inc.,  Stroudsburg.  Pp.  283-306. 

webb,  s.  D.  1969.  The  Burge  and  Minnechaduza  Clarendonian  mammalian  faunas  of  north-central  Nebraska. 
Univ.  California  Pub.  Dept.  Geol.  Sci.  78,  1-191. 

woodburne,  m.  o.  and  bernor,  r.  l.  1980.  On  superspecific  groups  of  some  Old  World  hipparionine  horses. 
J.  Paleont.  in  press. 


BRUCE  J.  MACFADDEN 

Florida  State  Museum 
University  of  Florida 
Gainesville,  Florida  32611 
U.S.A. 


Typescript  received  25  July  1979 


THE  TOARCI AN  AGE  OF  THE  UPPER  PART  OF 
THE  MARLSTONE  ROCK  BED  OF  ENGLAND 

by  M.  K.  HOWARTH 


Abstract.  The  ‘Transition  Bed’  of  Oxfordshire,  Northamptonshire,  and  Leicestershire  is  the  weathered  or 
altered  top  of  the  Marlstone  Rock  Bed.  In  the  top  0-05  m-0-3  m of  the  bed,  the  green  ferrous  minerals  were 
oxidized  to  limonite,  partly  before  deposition  of  overlying  beds,  partly  recently  in  some  areas.  In  another  type  of 
alteration,  best  seen  at  Harston,  Leicestershire,  much  granular  iron-pyrites  was  deposited  in  a highly  irregular 
zone  up  to  008  m thick  at  the  top  of  the  bed.  In  these  Midland  counties  the  whole  of  the  Tenuicostatum  Zone,  the 
basal  zone  of  the  Toarcian,  is  represented  in  the  top  1 -3  m of  the  Marlstone  Rock  Bed,  the  lower  3-6  m of  which 
belongs  to  the  Spinatum  Zone.  Regardless  of  the  depth  of  weathering  or  alteration,  Tiltoniceras  antiquum  and 
Dactylioceras  semicelatum  of  the  Semicelatum  Subzone  occur  widely  in  the  top  0T  m of  the  bed,  D.  tenuicostatum 
occurs  more  locally  at  a slightly  lower  horizon,  and  lower  still  one  D.  crosbeyi  is  evidence  for  the  Clevelandicum 
Subzone.  Ammonites  from  the  Semicelatum,  Tenuicostatum,  and  Paltum  Subzones  occur  in  the  Dorset  coast 
Marlstone  Rock  Bed.  North  of  Lincoln  the  top  of  the  Bed  is  at  about  the  top  of  the  Spinatum  Zone,  while  the 
Tenuicostatum  Zone  is  divided  between  an  overlying  hard  mudstone  and  higher  grey  shales.  The  change  from 
the  underlying  ironstone/limestone  facies  to  the  overlying  clays/shales-with-nodules  facies  took  place  at  the  top 
of  the  Spinatum  Zone  in  Yorkshire,  but  at  the  top  of  the  Tenuicostatum  Zone  in  the  Midland  counties. 

T he  Marlstone  Rock  Bed  is  one  of  the  most  distinctive  lithological  horizons  in  the  English  Lias.  It  is 
typically  an  oolitic  limestone,  though  the  sand  content  becomes  significant  in  a few  places,  and  the 
ferrous  iron  content  is  high  enough  in  two  areas  for  it  to  be  used  as  an  iron-ore.  The  outcrop  (text- 
fig.  1),  known  from  numerous  building-stone  and  iron-ore  quarries  in  the  past,  extends  from  the 
Dorset  coast  generally  north-eastwards  to  north  of  the  Humber.  Although  it  is  represented 
immediately  north  of  the  Mendips  and  also  at  Dundry,  south  of  Bristol,  it  is  generally  absent  around 
and  to  the  south  of  Bath.  There  is  another  gap  in  north  Northamptonshire  and  south  Leicestershire, 
where  it  is  absent  altogether  or  only  about  0-3  m thick,  owing  to  a combination  of  thin  deposition  and 
subsequent  erosion.  For  several  miles  north  and  south  of  Lincoln  it  disappears  and  is  represented  by  a 
layer  of  phosphatic  pebbles  and  possibly  some  thin  overlying  shales.  North  of  the  Humber  it  thins  out 
against  the  Market  Weighton  block,  and  it  does  not  reappear  farther  north  in  Yorkshire,  where 
equivalent  beds  are  developed  in  a different  facies. 

Traditionally  the  top  of  the  Marlstone  Rock  Bed  was  the  junction  between  the  Middle  and  Upper 
Lias,  or  more  specifically  the  junction  between  the  Spinatum  and  Tenuicostatum  Zones  (Arkell  1933, 
pp.  153-159;  Whitehead  et  al.  1952,  pp.  97,  105,  144-150).  In  fact  the  Marlstone  Rock  Bed  was 
referred  to  a single  zone,  the  Spinatum  Zone,  because  in  some  areas  it  has  a rich  fauna  of  species  of 
Pleuroceras  (Howarth  1958,  pp.  ix-xi).  The  beds  above  the  Marlstone  Rock  Bed  are  clays  and  shales 
in  most  areas,  but  in  some  parts  of  the  Midlands  the  ‘Transition  Bed’,  a bed  of  oolitic  limestone  up  to 
015  m thick,  is  the  immediately  overlying  bed.  It  was  first  described  from  the  Banbury  area,  north 
Oxfordshire,  and  later  at  Tilton,  Leicestershire,  and  contains  a rich  fauna  of  the  Upper  Lias 
ammonites  Dactylioceras  and  Tiltoniceras  ‘ acutum ’ (Blake).  It  was  called  the  Acutum  Zone  or 
hemera  by  Buckman  (19106,  p.  86)  and  the  Acutum  Subzone  by  Arkell  (1933,  p.  179),  and  placed  at 
the  base  of  the  Upper  Lias,  below  the  Tenuicostatum  Subzone,  as  the  lower  of  the  two  subzones  of  the 
Tenuicostatum  Zone.  This  subzonal  position  of  the  Transition  Bed  has  not  been  challenged  until 
recently,  though  Spath  ( 1 942,  p.  265;  1 956,  p.  1 43)  did  not  accept  the  validity  of  this  subdivision  of  the 
Tenuicostatum  Zone.  The  identification  of  the  horizon  to  which  the  Transition  Bed  belonged  was 
made  more  difficult  by  the  naming  of  an  atypical  representative  of  its  Dactylioceras  fauna  as 


IPalaeontology,  Vol.  23,  Part  3,  1980,  pp.  637-656,  pis.  80-82.| 


638 


PALAEONTOLOGY,  VOLUME  23 


Orthodactylites  directus  Buckman  (1926a,  pi.  654),  and  also  by  the  inability  of  anyone  to  find  either 
this  ammonite  or  Tiltoniceras  in  the  Yorkshire  coast  Upper  Lias  succession.  In  a more  recent 
investigation  of  the  succession  at  Tilton,  Hallam  (1955,  p.  21)  discovered  that  D.  directum  occurred  in 
the  top  0-9  m of  the  Marlstone  Rock  Bed  as  well  as  in  the  Transition  Bed,  and  suggested  that  it  would 
be  necessary  to  place  the  base  of  the  Toarcian  (i.e.  the  Upper  Lias)  at  least  0-9  m below  the  top  of  the 
Marlstone  Rock  Bed.  A more  conservative  view  was  taken  by  Howarth  (1958,  p.  xi),  and  followed 


text-fig.  1 . Sketch  map  of  the  outcrop  of  the  Marlstone  Rock  Bed,  showing  the  principal  localities  described  in 

the  text. 


later  by  Hallam  ( 1 967,  p.  397),  that  it  was  best  to  retain  the  Middle/Upper  Lias  boundary  at  the  top  of 
the  bed,  because  the  relationships  between  the  last  Pleuroceras  and  the  first  Dactylioceras  in  Britain 
were  not  known  at  that  time. 

No  further  advance  could  be  made  until  the  succession  of  ammonites  in  the  Tenuicostatum  Zone  of 
the  Yorkshire  coast  was  worked  out.  When  this  was  done  (Howarth  1 973,  enlarging  on  the  collecting 
of  the  late  Professor  P.  C.  Sylvester-Bradley,  whose  preliminary  results  were  published  in  Dean, 


HOWARTH:  AGE  OF  MARLSTONE  ROCK  BED 


639 


Donovan,  and  Howarth  1961,  p.  476)  the  following  sequence  of  ammonites,  and  of  subzones 
derived  from  them,  was  established: 


Zone 

Subzone 

Ammonite  faunas 

Dactylioceras 

semicelatum 

Tiltoniceras  antiquum  and 
Dactylioceras  semicelatum 

D.  semicelatum 

Dactylioceras 

D.  tenuicostatum 

D.  tenuicostatum 

tenuicostatum 

D.  clevelandicum 

D.  clevelandicum 

D.  crosbeyi 

Protogrammoceras 
pal  turn 

Protogrammoceras  paltum 

Abundant  faunas  of  Tiltoniceras  were  found  at  the  top  of  the  Tenuicostatum  Zone  in  Yorkshire,  not 
at  the  base  of  the  zone  where  the  genus  had  always  been  expected  before  (e.g.  Hallam  1967,  p.  415). 
This  alone  was  sufficient  to  suggest  that  the  Transition  Bed  of  the  Midlands  belonged  to  the  top 
subzone  of  the  Tenuicostatum  Zone,  and  confirmation  of  this  correlation  was  obtained  when  it  was 
found  that  most  of  the  Dactylioceras  in  that  bed,  to  which  the  name  D.  directum  had  always  been 
given  before,  were  typical  examples  of  D.  semicelatum.  In  fact  the  populations  of  the  latter  species  in 
the  Transition  Bed  and  in  the  Semicelatum  Subzone  of  the  Yorkshire  coast  are  very  similar,  having 
almost  identical  ranges  of  variation.  One  end  of  the  variation  consists  of  evolute  specimens  with  fine 
rectiradiate  ribs,  and  Buckman  gave  the  name  D.  directum  to  the  most  extreme  example  of  this  type 
from  the  Transition  Bed. 

Shortly  after  the  Yorkshire  coast  Tenuicostatum  Zone  succession  had  been  described,  an 
abundant  ammonite  fauna  was  found  in  the  top  of  the  Marlstone  Rock  Bed  in  a quarry  at  Harston, 
north  Leicestershire.  The  top  0 08  m of  the  bed  contained  many  D.  semicelatum  and  a few 
Tiltoniceras , and  was  clearly  equivalent  to  the  Transition  Bed,  though  it  was  not  developed  as  a 
distinct  bed  at  Harston.  The  main  discovery,  however,  was  the  presence  of  D.  tenuicostatum  in 
abundance  in  the  next  0-05  m below,  an  ammonite  that  had  hardly  ever  been  found  in  the  Marlstone 
Rock  Bed  before.  This  proved  the  presence  of  the  Tenuicostatum  Subzone  in  the  bed,  and  further 
minor  discoveries  showed  that  the  Clevelandicum  Subzone  occurred  lower  still  in  the  bed. 

The  presence  of  the  whole  of  the  Tenuicostatum  Zone  in  the  Marlstone  Rock  Bed  at  Harston,  and 
the  discovery  in  existing  museum  collections  of  specimens  of  Tiltoniceras  from  Tilton  preserved  in 
green  oolitic  limestone  typical  of  the  Marlstone  Rock  Bed,  led  to  further  investigation  of  the  Tilton 
Railway  Cutting.  It  was  found  that,  just  as  had  been  originally  described  by  Wilson  and  Crick  ( 1 889), 
the  Transition  Bed  is  not  a lithologically  distinct  bed,  it  is  merely  the  weathered  top  of  the  Marlstone 
Rock  Bed.  There  is  no  lithological  break  or  disconformity  that  marks  off  a distinct  bed  at  the  top, 
only  a highly  irregular  zone  of  oxidation  of  the  green  ferrous-iron  content  of  the  oolite  to  brown 
limonite.  The  Semicelatum  Subzone  ammonite  fauna  occurs  in  the  top  0-9  m of  this  complete 
Marlstone  Rock  Bed  (i.e.  including  the  ‘Transition  Bed’)  at  Tilton.  Unfortunately  there  is  no 
evidence  for  lower  subzones  of  the  Tenuicostatum  Zone  at  Tilton,  though  there  is  plenty  of  room  for 
them  above  the  highest  recorded  Pleuroceras  at  about  3 m below  the  top  of  the  Marlstone  Rock  Bed. 

The  ‘Transition  Bed’  and  its  distinct  ammonite  fauna  is  also  well  developed  in  the  Banbury-Byfield 
area  of  north  Oxfordshire  and  west  Northamptonshire.  Although  more  constant  in  thickness,  it 
appears  possible  to  interpret  it  similarly  in  that  area  as  the  weathered  top  of  the  Marlstone  Rock  Bed, 
weathering  that  probably  occurred  before  deposition  of  any  overlying  beds.  Between  north 


640 


PALAEONTOLOGY,  VOLUME  23 


Oxfordshire  and  south  Somerset  Tenuicostatum  Zone  ammonites  are  rarely  found  and  the  presence 
of  the  zone  within  the  Marlstone  Rock  Bed  has  yet  to  be  demonstrated.  On  the  Dorset  coast, 
however,  where  the  Marlstone  Rock  Bed  is  very  thin  (0-0-6  m),  ammonites  are  frequent  and  prove  the 
presence  of  the  Paltum,  Tenuicostatum,  and  Semicelatum  Subzones.  The  new  stratigraphical  work 
and  ammonite  collections,  and  reinterpretation  of  older  collections  are  described  in  detail  below. 

STRATIGRAPHICAL  DESCRIPTIONS 

1 . Dorset  coast.  The  Marlstone  Rock  Bed  forms  the  lowest  part  of  the  Middle  and  Upper  Liassic  Junction  Bed  in 
the  cliffs  between  Seatown  and  Eype,  and  has  been  described  in  detail  by  Buckman  (19226),  Jackson  (1922, 
1926),  and  Howarth  (1957).  The  bed  is  never  more  than  0-6  m thick  and  consists  of  three  layers,  the  lithological 
differences  and  ammonite  faunas  of  which  were  discussed  by  Howarth  (1957,  pp.  192-193).  The  lowest  layer  R is 
a coarse  conglomeratic  and  oolitic  limestone  that  contains  many  Pleuroceras  indicative  of  the  Apyrenum 
Subzone  of  the  Spinatum  Zone.  The  middle  layer  Px  is  a hard  grey  and  pink  limestone  with  scattered  ooliths  that 
contains  only  a few  P.  cf.  spinatum  and  probably  belongs  to  the  Hawskerense  Subzone.  The  top  layer  P is  a 
brown  finely  oolitic  limestone  that  contains  a rich  ammonite  fauna.  Previously  (Howarth  1957,  p.  193)  it  was  said 
to  be  of  Hawskerense  Subzone  age  only,  but  now  that  the  sequence  within  the  Tenuicostatum  Zone  is  known  in 
Yorkshire,  it  is  clear  that  layer  P is  a highly  condensed  bed  that  contains  most  horizons  from  the  Hawskerense  up 
to  the  Semicelatum  Subzones.  The  following  is  a list  of  the  ammonites  that  have  been  collected  from  layer  P: 

Dactylioceras  semicelatum  (BM  C.17548,  C.74719;  IGS  GSM  22475,  22514;  SM  J.44225-44226;  NMW  26.135 
G123),  D.  tenuicostatum  (NMW  26.135  G5-8  (9  specimens),  G124),  Protogrammoceras  paltum  (BM  67939, 
C.2200,  C.30769,  C.68536;  IGS  GSM  47160-47161,  49291;  SM  J.44789),  Pleuroceras  spinatum  (Bruguiere), 
P.  spinatum  var.  buckmani  (Moxon),  P.  yeovilense  Howarth,  P.  hawskerense  (Y oung  and  Bird),  P.  apyrenum 
(Buckman). 

These  ammonites  are  characteristic  of  the  Semicelatum,  Tenuicostatum,  Paltum,  and  Hawskerense  Subzones, 
and  the  only  horizon  for  which  there  is  no  evidence  is  the  Clevelandicum  Subzone.  The  examples  of  D.  semi- 
celatum (PI.  8 1 , figs.  3, 4,  and  Howarth  1 957,  pi.  1 7,  figs.  5, 6)  are  typical  of  the  species  and  match  Y orkshire  coast 
examples  closely.  The  ten  specimens  of  D.  tenuicostatum  (PI.  82,  figs.  5-8)  are  small  and  very  similar  to  examples 
from  near  the  top  of  the  Marlstone  Rock  bed  at  Harston,  Leicestershire;  they  all  came  from  a layer  of  fine  brown 
oolite  which  also  contained  one  specimen  of  D.  semicelatum,  many  gastropods,  and  a unique  Terebratulid  that 
was  later  described  as  ‘ Terebratula ’ reversa  Ager  (1956a,  p.  4,  pi.  1,  fig.  6)  (possibly  a Lobothyris).  This 
association  of  fossils  found  in  only  a single  block  was  the  basis  for  the  proposal  of  the  layer  At  by  Jackson  (1926, 
p.  497)  (At  was  derived  from  Buckman’s  hemera  ‘ athleticum  , a term  used  for  the  Transition  Bed  of  the  Midlands 
that  was  said  to  contain  a similar  Terebratulid).  However,  the  lithology  is  not  different  from  layer  P,  the 
brachiopod  has  no  special  age  significance,  and  the  ammonites  are  intermediate  in  age  between  the  Semicelatum 
and  Paltum  Subzones  ammonites  that  are  found  in  many  other  blocks  of  layer  P.  Therefore,  there  is  no 
justification  for  the  recognition  of  a separate  At  layer.  Well-preserved  specimens  of  Protogrammoceras  paltum  in 
layer  P include  the  holotype  and  paratype  (Buckman  1922a,  pi.  362A;  1923a,  pi.  362B),  an  example  figured  by 
Wright  (1884,  pi.  81 , figs.  4-6),  and  the  specific  synonym  Platyharpites platypleurus  Buckman  (1927a,  pi.  698).  So 
layer  P,  though  never  more  than  0-3  m thick,  contains  highly  condensed  representatives  from  the  Hawskerense  to 
Semicelatum  Subzones.  The  Marlstone  Rock  Bed  of  the  Dorset  coast,  i.e.  layers  R,  Px,  and  P,  belongs  to  the 
whole  of  the  Spinatum  and  Tenuicostatum  Zones,  so  it  is  approximately  equally  divided  between  the  Middle  and 
Upper  Lias.  The  next  higher  blocks  of  the  Junction  Bed  are  the  layers  N,  O,  and  D,  which  are  lateral  equivalents 
of  each  other,  and  contain  specimens  of  Harpoceras  exaratum,  from  about  the  middle  of  the  Exaratum  Subzone. 
There  are  no  records  of  Eleganticeras  that  would  indicate  the  presence  of  the  lower  part  of  the  Exaratum 
Subzone. 

2.  North  Dorset,  Somerset,  Avon,  and  Gloucestershire.  Northwards  from  the  Dorset  coast  the  Marlstone  Rock 
Bed  thickens  quickly,  and  the  term  Junction  Bed  is  now  applied  to  the  overlying  sequence  of  clays  and  limestones 
of  the  Upper  Lias.  Both  beds  are  very  rich  in  ammonites  in  the  Ilminster  area  of  south  Somerset.  In  the  well- 
known  Barrington  succession  described  by  Hamlet  ( 1 922),  Spath  ( 1 922),  and  Pringle  and  Templeman  ( 1 922),  the 
Marlstone  Rock  Bed  contains  many  Pleuroceras,  and  is  overlain  by  bed  1 (of  Hamlet),  a 0-175  m bed  of ‘sandy 
marl’  which  contains  D.  cf.  tenuicostatum  in  addition  to  more  examples  of  Pleuroceras.  Bed  2,  a 0- 1 m bed  of  grey 
oolitic  limestone,  contains  D.  semicelatum,  of  which  an  example  is  figured  here  (PI.  8 1 , figs.  1 , 2).  The  overlying 
bed  3 is  clay  containing  argillaceous  limestone  nodules,  and  is  of  Exaratum  Subzone  age.  So  the  Tenuicostatum 


HOWARTH:  AGE  OF  MARLSTONE  ROCK  BED 


641 


Zone  is  confined  to  bed  2 and  part  of  bed  1 , and  these  may  be  a local  lithological  variation  of  the  Marlstone 
Rock  Bed. 

The  Marlstone  Rock  Bed  is  well  developed  around  Batcombe  and  Evercreech,  near  Shepton  Mallet  on  the 
south  side  of  the  Mendips  (Richardson  1906, 1909),  but  evidence  for  the  presence  of  the  Tenuicostatum  Zone  has 
not  been  obtained.  After  a gap  north  of  the  Mendips,  the  bed  reappears  north  of  Bath,  thickens  quickly,  and  was 
formerly  extensively  quarried  along  the  western  escarpment  of  the  Cotswolds  in  Gloucestershire.  There  is  little 
ammonite  evidence  for  the  age  of  the  top  of  the  bed.  Species  of  Pleuroceras  from  both  subzones  of  the  Spinatum 
Zone  are  common  at  some  localities  (e.g.  Alderton  Hill),  but  data  about  their  stratigraphical  position  in  the 
Marlstone  Rock  Bed  are  lacking.  There  are  no  Tenuicostatum  Zone  ammonites  in  existing  museum  collections 
from  this  area.  One  record  is  intriguing,  however:  in  an  exposure  of  the  Marlstone  Rock  Bed  near  Stow-on-the- 
Wold,  about  25  km  east  of  the  Cotswolds  escarpment,  Hull  (1857,  pp.  19, 20)  saw  a ‘band  of  deep  reddish  purple 
ironstone’  015  m thick  at  the  top  of  the  bed  ‘filled  with  good  specimens  of  Ammonites  annulatus ’.  It  is  likely  that 
these  were  examples  of  D.  tenuicostatum  or  D.  semicelatum  and  they  would  show  that  most  of  the  Tenuicostatum 
Zone  was  in  the  Rock  Bed.  The  exposure  was  not  extant  in  1929  when  Richardson  (1929,  p.  31)  quoted  the 
record,  and  the  ammonites  are  not  preserved  in  the  Institute  of  Geological  Sciences,  so  the  occurrence  cannot  be 
investigated  further.  In  the  Stowell  Park  bore-hole,  1 8 km  south-west  of  Stow-on-the-Wold,  the  Marlstone  Rock 
Bed  did  not  yield  any  ammonites,  but  1 m of  overlying  shales  contained  Tiltoniceras  and  Dactylioceras  of  the 
Semicelatum  Subzone.  This  shows  that  at  least  some  of  the  Tenuicostatum  Zone  is  above  the  Marlstone  Rock 
Bed,  though  it  need  not  be  more  than  the  upper  half  of  the  Semicelatum  Subzone. 

3.  Oxfordshire  and  Northamptonshire.  The  Marlstone  Rock  Bed  used  to  be  extensively  quarried  for  iron-ore  and 
building  stone  over  a large  area  between  Banbury  and  Northampton,  and  details  of  the  many  former  quarries 
can  be  found  in  Whitehead  et  al.  (1952).  It  was  in  this  district  that  the  term  ‘Transition  Bed’  was  first  proposed  by 
Walford  (1878,  p.  2)  for  a pale-brown  oolitic  and  ferruginous  ‘marl’  0 050-0-075  m thick  that  forms  the  top  of 
the  Marlstone  Rock  Bed.  The  type  area  is  around  Banbury,  and  the  best-known  localities  were  quarries  at 
Adderbury,  King’s  Sutton,  and  Middleton  Cheney,  south  and  east  of  Banbury.  Large  numbers  of  T.  antiquum 
(Wright)  and  D.  semicelatum  (including  the  holotype  of  D.  ‘ directum ’ Buckman)  and  many  small  gastropods 
were  obtained  from  the  Transition  Bed  in  these  quarries,  and  they  show  that  the  bed  belongs  to  the  Semicelatum 
Subzone.  Arkell  (1947,  p.  21)  proposed  that  the  term  ‘Acutum  Bed’  (after  Tiltoniceras  ''acutum'  Blake,  the 
holotype  of  which  came  from  Adderbury)  should  supercede  Transition  Bed  in  the  north  Oxfordshire  area,  but 
this  change  of  name  has  not  been  adopted  by  other  authors  (‘Acutus  Subzone’  had  been  used  earlier  by  Walford 
(1899,  p.  33)).  This  lithology  and  0-050-0-075  m thickness  is  fairly  constant  over  the  whole  of  north  Oxfordshire 
and  west  Northamptonshire  as  far  north  as  Daventry.  At  Iron  Cross  Farm,  Byfield,  the  last  locality  at  which  it 
was  well  exposed  (Howarth  1978,  p.  240),  it  forms  the  upwards  continuation  of  the  Marlstone  Rock  Bed,  with 
which  it  has  a sharp  and  irregular  junction.  The  Transition  Bed  appears  to  be  the  altered  top  of  the  Marlstone 
Rock  Bed,  alteration  that  is  mainly  oxidation  and  leaching  of  the  green  ferrous  iron,  and  which  probably  took 
place  before  deposition  of  the„overlying  Abnormal  Fish  Bed.  The  latter  is  separated  by  a parting  from  the  top  of 
the  Transition  Bed,  and  is  of  mid  and  upper  Exaratum  Subzone  age  (Howarth  1978,  p.  241). 

In  areas  further  east,  and  especially  around  Milton  and  Bugbrooke  west  of  Northampton,  a series  of  beds  up 
to  0-35  m thick  has  been  referred  to  the  Transition  Bed  (Thompson  1889, 1892).  This  is  due  to  the  inclusion  of  an 
overlying  sandy  or  shaly  clay  that  does  not  contain  the  characteristic  Transition  Bed  ammonites  or  gastropods. 
The  age  of  the  clay  is  not  accurately  known,  but  it  may  bridge  the  small  disconformity  that  occurs  everywhere 
else  between  the  Transition  Bed  and  the  Abnormal  Fish  Bed,  and  it  should  not  be  included  in  the  Transition  Bed. 
At  Bugbrooke  Thompson  (1892,  p.  337)  said  that  the  Transition  Bed  was  not  present  as  a distinct  bed,  but  was 
nevertheless  clearly  shown  by  the  altered  character  of  the  top  of  the  Marlstone  Rock  Bed  which  contained  the 
Transition  Bed  fossils.  Thus  it  appears  that  throughout  the  area  the  Transition  Bed  is  the  altered  top  of  the 
Marlstone  Rock  Bed,  alteration  which  probably  took  place  before  deposition  of  the  overlying  beds. 

Evidence  for  the  subzonal  age  of  the  remainder  of  the  Marlstone  Rock  Bed  is  meagre  in  this  area.  Only  one 
specimen  of  D.  tenuicostatum  has  been  found  (PI.  82,  figs.  3,  4),  at  Rothersthorpe,  5 km  south-west  of 
Northampton,  from  an  unrecorded  horizon,  but  judging  from  the  grey-green  finely  oolitic  matrix,  probably 
from  immediately  below  the  altered  top  of  the  Marlstone  Rock  Bed.  This  ammonite  is  evidence  for  the 
Tenuicostatum  Subzone,  but  there  are  no  ammonites  to  prove  the  presence  of  lower  subzones.  The  majority  of 
the  Marlstone  Rock  Bed  belongs  to  the  Spinatum  Zone  and  the  characteristic  brachiopods  Tetrarhynchia 
tetrahedra  and  Lobothyris  punctata  are  abundant  except  at  the  extreme  top.  Pleuroceras  is  rare  in 
Northamptonshire:  a few  P.  spinatum  have  been  found  and  at  least  one  P.  apyrenum  is  known,  but  their  horizons 
are  not  recorded.  The  Middle/Upper  Lias  junction  occurs  near  the  top  of  the  Marlstone  Rock  Bed,  probably 
within  the  top  0-25  m. 


642 


PALAEONTOLOGY,  VOLUME  23 


4.  Tilton,  Leicestershire.  The  well-known  section  at  Tilton  Railway  Cutting  (SK  762055)  was  first  described  by 
Wilson  and  Crick  (1889)  and  there  is  a good  photograph  of  it  in  its  original  state  in  Fox-Strangways  (1903,  p.  30, 
pi.  2).  Wilson  and  Crick  saw  the  Marlstone  Rock  Bed  soon  after  it  was  uncovered  below  a thickness  of  up  to  9 m 
of  Upper  Lias  shales  and  it  had  been  little  affected  by  recent  subaerial  weathering.  The  ‘Transition  Bed’  was 
described  as  a flaggy  limestone  0T  5-0-23  m thick  containing  a distinctive  fauna,  especially  the  ammonite 
Tiltoniceras,  even  though  they  said  that  ‘it  possesses  the  mineral  characters  of  and  is  welded  to  the  top  of  the 
Marlstone  Rock  Bed’  (Wilson  and  Crick  1889,  p.  297).  Woodward  (1893,  p.  236)  repeated  this  interpretation  of 
the  Transition  Bed,  but  Whitehead  et  al.  (1952,  p.  135)  made  no  mention  of  a Transition  Bed  at  Tilton,  nor 
anywhere  in  the  surrounding  area.  The  railway  section  was  again  described  by  Hallam  (1955;  1968,  p.  208)  who 
recognized  the  01 5-0-23  m Transition  Bed,  and  observed  that  it  lapped  ‘over  minor  irregularities  at  the  top  of 
the  ironstone’  and  ‘rested  non-sequentially  on  the  ironstone’.  The  lithology  of  the  Transition  Bed  has  been 
described  as  a pale-brown  or  cream  finely  oolitic  limestone,  sometimes  flaggy,  and  sometimes  passing  up  into 
sandy  marl.  Tiltoniceras  preserved  in  such  brown  oolitic  limestone  is  very  common,  but  many  others  also  occur 
preserved  in  the  deep-green  oolitic  ironstone  that  is  typical  of  the  Marlstone  Rock  Bed  at  Tilton.  Hallam  (1955, 
p.  21)  explained  the  latter  by  saying  that  the  genus  occurred  rarely  in  the  ironstone  immediately  below  the 
Transition  Bed. 

Examination  of  the  Tilton  Railway  Cutting  exposures  in  recent  years  shows  that  the  Transition  Bed  does  not 
exist  as  a separate  bed.  It  is  the  weathered  top  of  the  Marlstone  Rock  Bed,  in  which  the  siderite  and  chamosite  of 
the  deep  green  oolitic  limestone  are  oxidized  to  limonite;  partial  decalcification  gives  it  a friable,  granular 
texture,  which  has  been  described  as  sandy,  though  the  bed  is  not  arenaceous.  The  depth  of  weathering  varies 
greatly  between  0-01  m and  0-25  m below  the  top  surface,  and  the  lowest  extent  is  marked  by  an  undulating  thin 
sheet  of  brown  limonite.  Many  specimens  of  T.  antiquum  (Wright),  D.  semicelatum  (PI.  81,  figs.  5,  6)  and 
Gibbirhynchia  tiltonensis  Ager,  and  many  small  gastropods  (Wilson  and  Crick  1889,  pp.  298-305,  pi.  9)  occur  in 
the  top  0-2  m of  the  Marlstone  Rock  Bed.  Whether  they  occur  preserved  in  deep-green  ironstone  or  pale-brown 
oolitic  limestone  depends  entirely  on  how  deeply  the  weathering  has  penetrated  at  any  particular  point.  Several 
fine  examples  of  Tiltoniceras  preserved  in  green  ironstone  were  obtained  from  only  0-025  m below  the  top,  but 
most  of  the  green  specimens  occur  lower  down.  In  some  specimens  that  are  orientated  approximately  vertically 
in  the  bed,  the  upper  half  of  the  ammonite  is  pale  brown  and  the  lower  half  deep  green.  Weathering  also 
penetrates  deeply  down  some  of  the  vertical  joints  and  can  convert  fossils  much  lower  down  into  pale-brown 
friable  limestone.  In  a few  places  horizontal  bedding  planes  lead  to  greater  penetration  of  weathering,  and  rarely 
the  whole  of  the  top  0-2  m is  affected  giving  the  appearance  of  a distinct  lithological  bed  at  the  top  of  the  iron- 
stone. Such  beds  fade  out  rapidly  laterally,  and  the  usual  state  is  dark-green  Marlstone  Rock  Bed  weathered 
brown  to  a highly  variable  depth. 

D.  semicelatum  is  commonest  in  the  top  0-2  m,  but  unlike  Tiltoniceras  it  also  occurs  lower  down  to  depths  of 
0-9  m below  the  top  of  the  ironstone.  This  is  the  amount  of  the  Marlstone  Rock  Bed  that  must  be  referred  to  the 
Semicelatum  Subzone.  No  Upper  Lias  ammonites  belonging  to  lower  subzones  occur  at  Tilton.  The  only 
Pleuroceras  found  in  situ  is  a specimen  of  P.  cf.  hawskerense  (Young  and  Bird)  3-0  m below  the  top  of  the 
Marlstone  Rock  Bed,  and  it  indicates  the  Hawskerense  Subzone  of  the  Spinatum  Zone.  The  brachiopods 
Tetrarhynchia  tetrahedra  (J.  Sowerby)  and  L.  punctata  (J.  Sowerby)  range  higher  in  the  ironstone,  the  last  ones 
being  about  1-2  m below  the  top  (Hallam  1955,  p.  20).  These  two  are  usually  held  to  be  good  indicators  of  the 
Spinatum  Zone  in  England,  but  there  are  rare  records  from  the  Upper  Lias,  the  genus  Tetrarhynchia  ranges  up 
into  the  Bajocian  (Ager  19566,  p.  3),  and  T.  tetrahedra  occurs  in  the  Upper  Lias,  Bifrons  Zone,  in  Spain  (Hallam 
1972,  p.  408).  So  in  the  absence  of  ammonites,  it  does  not  seem  safe  to  take  the  highest  occurrence  of  these 
brachiopods  as  unequivocal  evidence  of  the  Spinatum  Zone.  The  evidence  available  at  present  suggests  that  the 
Spinatum/Tenuicostatum  Zone  boundary  occurs  between  1 m and  3 m below  the  top  of  the  Marlstone  Rock  Bed 
at  Tilton.  There  is  room  in  this  thickness  for  condensed  representatives  of  the  three  lower  subzones  of  the 
Tenuicostatum  Zone,  and  a disconformity  need  not  be  postulated  to  explain  their  absence.  There  is  also  no 
lithological  evidence  for  such  a disconformity. 

The  beds  above  the  Marlstone  Rock  Bed  are  clays  and  shales  with  a few  thin  beds  of  limestone  or  limestone 
nodules.  The  basal  2-8  m belongs  to  the  Exaratum  Subzone,  and  uncrushed  examples  of  Harpoceras  elegans 
(J.  Sowerby)  and  H.  serpentinum  (Schlotheim)  occur  at  about  the  2 m level.  These  indicate  the  top  part  of  the 
Exaratum  Subzone,  and  the  absence  of  H.  exaratum  suggests  that  the  non-sequence  between  the  top  of  the 
Marlstone  Rock  Bed  and  the  shales  represents  at  least  the  lower  half  of  the  Exaratum  Subzone.  All  the  higher 
shales  up  to  the  top  of  the  cutting  belong  to  the  Falciferum  Subzone  and  contain  the  index  ammonite  commonly 
throughout.  The  following  is  a summary  of  the  section  exposed  in  the  Tilton  Railway  Cutting  (SK  762055), 
Leicestershire: 


HOWARTH:  AGE  OF  MARLSTONE  ROCK  BED  643 

Zone  and  subzone  of  Harpoceras  falciferum 

Grey  shale,  with  two  rows  of  small  limestone  nodules  about  0-5  m and  0-6  m below  the  top.  H.  falci- 
ferum   5-50  m 

Grey  clay  containing  large  calcite  ooliths.  H. falciferum,  Phylloceras  heterophyllum  (J.  Sowerby)  0-70  m 

Subzone  of  Harpoceras  exaratum 

Grey  clay,  oolitic.  Large  specimens  of  H.  serpentinum  (Schlotheim) 0-80  m 

Grey  oolitic  limestone.  H.  elegans  (J.  Sowerby)  (BM  C.8048 1-80483)  and  H.  serpentinum  common 

(BM  C.80484-80485),  and  many  Dactylioceras  sp.  indet 0-20  m 

Grey  shales,  paper  shales,  and  clays.  H.  serpentinum  in  top  0-5  m l-80m 


Zones  of  Dactylioceras  tenuicostatum  and  Pleuroceras  spinatum 

Marlstone  Rock  Bed: 

a.  Ironstone.  Dark-green  finely  oolitic  limestone,  containing  chamosite  and  siderite,  weathered 
brown  to  an  irregular  depth,  and  sometimes  more  deeply  along  joints  and  bedding  planes. 
Tiltoniceras  antiquum  (Wright)  (BM  C.  10265-10267,  C.41733,  C.48753-48757,  C.80242-80276, 


C. 80470-80480)  and  D.  semicelatum  (BM  C.36186-36188,  C.49766,  C.80277-80282,  C.80466- 
80469)  are  abundant  in  the  top  0-2  m and  0-9  m respectively  and  indicate  the  Semicelatum 
Subzone;  P.  cf.  hawskerense  (Young  and  Bird)  (BM  C.73686)  occurs  at  the  bottom  and  indicates 
the  Hawskerense  Subzone 3-0  m 

b.  Green  oolitic  limestone,  containing  numerous  specimens  of  Tetrarhynchia  tetrahedra  and 

Lobothyris punctata,  and  many  bivalves  (band  B of  Hallam  1955,  p.  18) 0-45  m 

c.  Sandrock.  Green  massive  calcareous  sandstone  . . . 1 -4  m 

d.  Calcareous  sandstone  as  bed  c,  but  with  many  nests  of  the  brachiopods  T.  tetrahedra  and 

L.  punctata  {band  AofHallam  1955, p.  18) 0-3  m 

e.  Sandrock,  as  bed  c 0-75  m 


Other  exposures  of  the  Marlstone  Rock  Bed  in  the  Tilton  area  were  in  iron-ore  quarries,  where  the  bed  had 
been  less  deeply  buried  than  in  the  railway  cutting.  Consequently  the  top  of  the  bed  had  been  more  strongly 
weathered.  Of  those  described  by  Whitehead  et  al.  (1952)  and  Hallam  (1955,  1968),  few  now  remain.  One  that 
can  still  be  seen  is  the  old  quarry  1-3  km  east  of  Tilton  (SK  756056),  where  the  top  0-2  m of  the  Marlstone  Rock 
Bed  is  highly  weathered  into  a pale-brown  oolitic  limestone  that  contains  Tiltoniceras  antiquum  and 
Dactylioceras  semicelatum.  Other  quarries,  now  obscured,  were  similar,  and  it  is  thought  that  the  ‘Transition 
Bed’  is,  in  all  cases,  the  weathered  top  of  the  Marlstone  Rock  Bed. 

5.  Grantham  area,  north  Leicestershire  and  south  Lincolnshire.  Two  quarries  working  the  Marlstone  Rock  Bed  as 
iron-ore  existed,  until  closed  down  and  filled  in  in  1975,  at  Harston,  12  km  south-west  of  Grantham.  Here  the  top 
of  the  Marlstone  Rock  Bed  contains  more  Dactylioceratidae  than  any  other  exposure  of  the  bed  in  England,  and 
it  is  the  most  important  section  for  dating  the  Upper  Lias  part  of  the  bed.  A section  for  the  Upper  Lias  shales 
above  the  bed  was  given  in  Hallam  (1968,  p.  210),  but  a more  detailed  description  is  now  given,  so  that  the 
position  of  the  disconformities  can  be  established.  Section  at  Harston  Quarry  (SK  843305),  1 -5  km  south-south- 


east of  Harston: 

Zone  and  subzone  of  Harpoceras  falciferum 

Clay.  Impressions  of//,  falciferum  and  Dactylioceras  sp.  indet. . 2-00  m 

Brown  rubbly  limestone,  oolitic  or  pisolitic  in  places.  H.  falciferum  ......  0-20  m 

Subzone  of  Harpoceras  exaratum 

Grey  shale 1 -20  m 

Scattered  flat  nodules  of  blue  limestone,  weathered  red-brown  and  white.  H.  elegans  (J.  Sowerby) 
abundant,  Dactylioceras  anguiforme  Buckman  abundant,  Nodicoeloceras  crassoides  (Simpson), 
Phylloceras  heterophyllum  (J.  Sowerby),  Coelodiscus  minutus  (Scbub\er)  . . . . . OTOm 

Grey  shale c.  10-00  m 

Grey  calcareous  clay,  forming  a hard  massive  bed.  A few  limestone  nodules  occur  in  a row  at  the  top. 

Large  specimens  of  H.  elegans,  H.  serpentinum  (Schlotheim)  and  Hildaites  murleyi  (Moxon)  . l-30m 

Shale,  with  a row  of 0-025  m thick  flat  limestone  nodules  at  the  top.  H.  serpentinum,  H.  elegans  . . 0-05  m 

Scattered  lenticles  of  coarse  sandstone,  cross-bedded,  with  many  granules  of  iron  pyrites  and  some 
small  pebbles.  Much  shell  debris  broken  into  small  fragments.  Fragments  of  Harpoceras  (?//.  cf. 
exaratum) 0-0-05  m 


644  PALAEONTOLOGY,  VOLUME  23 

Zone  of  Dactylioceras  tenuicostatum 

Marlstone  Rock  Bed: 

a.  Pale-brown  limestone,  consisting  of  numerous  calcite  ooliths  and  minute  shell  fragments  in  a 

calcareous  matrix;  the  top  0-05  m contains  patches  of  crystalline  calcite  and  occasional  pebbles  of 
brown  limestone;  the  lower  half  becomes  green-coloured,  more  coarsely  oolitic,  with  chamosite 
and  siderite,  and  much  recrystallized  calcite;  the  top  0-025-0-080  m is  full  of  fine  granules  of  iron- 
pyrites  and  is  grey-green  in  colour;  its  very  uneven  lower  boundary  is  marked  by  a solid  line  of 
iron-pyrites,  and  the  bed  below  is  pale  brown  with  only  a few  granules  of  iron-pyrites  . . . 1 -20  m 

Subzone  of  Dactylioceras  semicelatum 

The  top  0-08  m contains  Tiltoniceras  antiquum  (Wright)  (BM  C.80237-80241),  D.  semicelatum 
common  (BM  C. 80169,  C.80171-80235),  Acrocoelites  vulgaris  (Young  and  Bird),  Gibbirhynchia 
sp.  and  gastropods. 

Subzone  of  Dactylioceras  tenuicostatum 

Between  0 08  m and  013  m below  the  top  D.  tenuicostatum  is  abundant  (BM  C.80099-80168)  and 
Gibbirhynchia  sp.  occurs. 

Subzone  of  Dactylioceras  clevelandicum 

0-23  m below  the  top  one  specimen  of  D.  crosbeyi  (Simpson)  (BM  C.80170)  was  found. 

Zone  of  Pleuroceras  spinatum 

b.  Deep-green  oolitic  limestone,  with  much  chamosite  and  siderite,  and  many  bands  of  recrystallized 
calcite.  Abundant  Tetrarhynchia  tetrahedra  and  Lobothyris  punctata  in  nests.  One  Pleuroceras  cf. 
spinatum  (Bruguiere)  0-25  m below  the  top,  and  several  other  specimens  not  in  situ 

300  m 

This  quarry  contained  one  of  the  best  sections  of  the  Marlstone  Rock  Bed  for  demonstrating  that  the  top  part 
that  contains  Tiltoniceras  is  a typical  part  of  the  bed  that  has  been  diagenetically  altered.  The  alteration  is  due  to 
pyritization  from  the  top  surface  downwards.  It  consisted  of  the  deposition  of  a large  amount  of  fine  granular 
iron-pyrites,  which  penetrated  to  a depth  varying  between  0-025  m and  0 080  m and  the  very  uneven  lower 
boundary  is  marked  by  a thin  sheet  of  solid  iron-pyrites.  Tiltoniceras  and  D.  semicelatum  (PI.  81,  figs.  10,  11; 
PI.  82,  figs.  11,12)  occur  in  the  top  0-080  m,  so  some  of  the  Semicelatum  Subzone  is  in  the  pyritized  part  and  some 
in  the  unaltered  part  below.  Most  ammonites  lie  parallel  to  the  bedding  plane,  but  a few  are  at  a high  angle  and 
occasionally  the  lower  boundary  of  pyritization  has  reached  half  down  an  ammonite.  There  is  no  lithological 
break  or  change  other  than  the  pyritization,  except  for  a few  pebbles  just  below  the  top  surface. 

D.  tenuicostatum  (PI.  82,  figs.  1,  2,  9,  10)  occurs  in  abundance  between  0-08  m and  0-13  m below  the  top  of  the 
Bed,  and  this  is  the  extent  of  the  Tenuicostatum  Subzone,  which  is  below  the  pyritized  zone.  A single  D.  crosbeyi 
0-23  m below  the  top  is  evidence  for  the  presence  of  the  Clevelandicum  Subzone.  There  are  no  ammonites  to 
prove  the  presence  of  the  Paltum  Subzone,  but  there  is  plenty  of  room  for  it  between  0-25  m and  1-2  m below  the 
top  of  the  Bed.  Marlstone  Rock  Bed  division  b of  the  above  section  is  a natural  downward  continuation  of  the 
upper  part  where  it  becomes  richer  in  iron,  and  several  specimens  of  Pleuroceras  occur  of  the  Spinatum  Zone. 

The  whole  of  the  Tenuicostatum  Zone  is  in  the  top  1 -2  m of  the  Marlstone  Rock  Bed  at  Harston,  and  there  are 
no  major  lithological  discontinuities  within  that  part  of  the  bed.  The  main  disconformity  is  at  the  top  of  the  bed 
where  the  lithology  changes  to  shale  facies,  and  the  bottom  one-third  of  the  Exaratum  Subzone  is  missing 
(because  of  the  absence  of  Eleganticeras).  The  middle  third  of  that  subzone  is  represented  only  by  the  lenticles  of 
sandstone  that  contain  Harpoceras,  and  continuous  deposition  starts  only  in  the  upper  third  of  the  subzone 
where  more  than  1 1 m of  shales  contain  H.  elegans  and  H.  serpentinum. 

At  Denton  Park  Quarry  (SK  856316),  1-5  km  north-east  of  Harston  Quarry,  a similar  succession  was  seen  at 
the  top  of  the  Marlstone  Rock  Bed,  though  ammonites  were  much  less  common.  The  top  0-03-008  m of  the  bed 
contains  much  granular  iron-pyrites  as  at  Harston,  but  it  is  more  of  a shell  bed  containing  a great  number  of 
broken  bivalve  shells  and  large  numbers  of  the  belemnite  Acrocoelites  vulgaris  (Young  and  Bird).  A few 
fragments  of  large  specimens  of  Tiltoniceras  were  also  seen.  The  lower  non-pyritized  part  of  the  bed  is  similar  to 
that  at  Harston,  but  Tenuicostatum  Subzone  ammonites  were  not  found. 

6.  North  Lincolnshire  and  south  Humberside.  North  of  Grantham  the  Marlstone  Rock  Bed  thins  steadily  and 
disappears  altogether  before  Lincoln.  At  Lincoln  the  Spinatum  Zone  is  absent  or  is  represented  by  a bed  of 


HOWARTH:  AGE  OF  MARLSTONE  ROCK  BED 


645 


phosphatic  pebbles  (Trueman  1918;  Howarth  1958,  p.  xii,  bed  11),  but  there  is  no  ammonite  evidence  for  the 
presence  of  the  zone  nor  for  the  lowest  three  subzones  of  the  Tenuicostatum  Zone.  The  presence  of  the 
Semicelatum  Subzone  is  shown,  however,  by  specimens  of  T.  antiquum  (Trueman  Coll.,  Nottingham  University, 
and  BM  C.48429-48432)  in  the  top  0-15  m of  the  0-60  m of  overlying  shales  (Howarth  1958,  p.  xi,  bed  12). 
D.  semicelatum  also  occurs  in  these  shales,  and  probably  in  the  shales  of  bed  14  above. 

The  Marlstone  Rock  Bed  reappears  north  of  Lincoln  and  it  was  well  exposed  in  recent  years  in  quarries  at 
Kirton  in  Lindsey  (Howarth  and  Rawson  1965)  and  Roxby  (Penny  and  Rawson  1969,  pp.  194-197).  The 
distribution  of  ammonites  in  the  Upper  Lias  shales  was  poorly  known  in  these  quarries,  and  better  records  have 
been  obtained  from  boreholes  in  the  same  area.  Most  information  came  from  boreholes  near  Worlaby,  8 km  east 
of  Roxby,  where  many  specimens  of  T.  antiquum  occurred  in  shales  between  4 m and  5-7  m above  the  Marlstone 
Rock  Bed  (Richardson  1979).  The  following  succession  for  part  of  the  Middle  and  Upper  Lias  in  this  area 
incorporates  details  of  the  Kirton  in  Lindsey  Quarry  (Howarth  and  Rawson  1965,  pp.  262-263),  the  north  end  of 
the  Roxby  Quarry  (Penny  and  Rawson  1969,  p.  196),  and  some  records  from  the  Worlaby  boreholes. 
Thicknesses  and  lithology  show  little  variation,  though  the  rows  of  doggers  are  more  obvious  in  the  quarries, 
especially  at  Kirton  in  Lindsey.  The  ammonite  distribution  is  the  same,  and  records  from  all  three  places  are 
included. 


Subzone  of  Harpoceras  exaratum 

Shale,  with  two  rows  of  doggers  and  a band  of  limestone.  Beds  25-29  at  Kirton  in  Lindsey;  bed  29, 
a row  of  doggers  0T3  m from  the  top,  contains  many  H.  elegans  (J.  Sowerby);  bed  27,  a bed  of 
limestone  1-2  m from  the  top,  contains  H.  cf.  exaratum  (Young  and  Bird);  bed  25  is  a row  of 


doggers  at  the  base 4-60  m 

Subzone  of  Dactylioceras  semicelatum 

Shale,  close-bedded,  but  sandy  in  basal  0-3  m.  Beds  23  and  24  at  Kirton  in  Lindsey.  Many  crushed 

Tiltoniceras  antiquum  (Wright),  sometimes  in  shell  beds,  through  most  of  the  thickness  . 3-50  m 

Shale,  with  scattered  limestone  nodules,  especially  near  the  base.  Crushed  D.  semicelatum  in  the 
shales  in  the  borehole,  and  well-preserved  solid  specimens  in  the  basal  nodules  at  Roxby  1 • 70  m 

Subzone  of  Dactylioceras  tenuicostatum 

Shale.  A few  D.  cf.  tenuicostatum 1 -30  m 


Subzones  of  Dactylioceras  tenuicostatum  (part),  D.  clevelandicum,  and  Protogrammoceras  paltum 
Hard,  pale-grey,  calcareous  mudstone,  silty  and  micaceous,  with  some  phosphatic  and  calcareous 
nodules.  Bed  21  at  Kirton  in  Lindsey.  Many  well-preserved  ammonites  and  belemnites: 

D.  tenuicostatum  and  D.  clevelandicum  common;  one  large  P.  paltum  (Buckman)  known  from 

Roxby 0-40-1  TOm 


Zone  of  Pleuroceras  spinatum 

Marlstone  Rock  Bed.  Green  oolitic  limestone.  Rare  Pleuroceras  cf.  hawskerense  (Young  and  Bird) 

(level  unknown).  Many  brachiopods 3 00  m 

The  Dactylioceratidae  that  were  recorded  previously  (Howarth  and  Rawson  1965,  p.  262)  in  bed  21  at  Kirton 
in  Lindsey  have  been  reassessed  in  the  light  of  the  succession  of  species  now  known  in  the  Yorkshire  coast  Grey 
Shales  (Howarth  1973).  Dactylioceras  tenuicostatum  and  D.  clevelandicum  are  both  present,  and  with  the  single 
large  Protogrammoceras  paltum  at  Roxby,  they  show  that  the  Paltum,  Clevelandicum,  and  part  of  the 
Tenuicostatum  Subzones  are  present  in  that  bed.  D.  tenuicostatum  also  occurs  in  the  shales  above,  and  then 
D.  semicelatum  and  Tiltoniceras  antiquum  occur  higher  up.  So  it  can  be  shown  that  the  Tenuicostatum  Zone  lies 
wholly  above  the  Marlstone  Rock  Bed  in  the  area  north  of  Lincoln.  Approximately  the  lower  half  of  the  zone  is 
condensed  in  a bed  up  to  I T m thick,  but  unlike  the  ‘Transition  Bed’  of  Oxfordshire  to  Leicestershire,  it  is  a 
calcified  silty  mudstone  significantly  different  in  lithology  from  the  Marlstone  Rock  Bed.  The  upper  half  of  the 
zone  occurs  in  shales  6-5  m thick  that  resemble  the  Grey  Shales  of  the  Yorkshire  coast  in  thickness  and  lithology, 
except  for  the  absence  of  pyritized  doggers. 


646 


PALAEONTOLOGY,  VOLUME  23 


SYSTEMATIC  DESCRIPTIONS 
Family  dactylioceratidae  Hyatt,  1867 
Genus  dactylioceras  Hyatt,  1867 
Subgenus  orthodactylites  Buckman,  1926 
Dactylioceras  ( Orthodactylites ) semicelatum  (Simpson) 

Plates  80,  81;  Plate  82,  figs.  11,  12;  text-figs.  2,  3 

1819  Ammonites  annulatus  J.  Sowerby,  p.  41,  pi.  222,  figs.  1,  2 ( non  figs.  3-5)  (non  Ammonites 
annulatus  Schlotheim,  1813). 

1843  Ammonites  semicelatus  Simpson,  p.  20. 

1855  Ammonites  semicelatus  Simpson,  p.  50. 

1884  Ammonites  semicelatus  Simpson,  p.  81. 

1911a  Dactylioceras  semicelatum  (Simpson);  Buckman,  pi.  31. 

1926a  Orthodactylites  directus  Buckman,  pi.  654. 

1927a  Kryptodactylites  semicelatus  (Simpson);  Buckman,  pi.  31A. 

1927a  Orthodactylites  mitis  Buckman,  pi.  738. 

1957  Dactylioceras  directum  (Buckman);  Howarth,  p.  197,  pi.  17,  figs.  5,  6. 

1957  Dactylioceras  semicelatum  (Simpson)  and  spp.;  Maubeuge,  figs.  1-3, 718-21, 41, 42, 44,  746, 47, 
48,749,  758,  759(1),  59  (2). 

1957  Dactylioceras  pseudocrassoides  Maubeuge,  p.  201,  pi.  13,  fig.  28. 

1957  Dactylioceras  densicostatum  Maubeuge,  p.  202,  pi.  13,  fig.  29. 

1960  Dactylioceras  sp.  indet.;  Hoffmann  and  Martin,  p.  114,  pi.  9,  fig.  5;  pi.  10,  figs.  2a,  2b. 

1968  Dactylioceras  cf.  toxophorum  (Buckman);  Hoffmann,  p.  4,  pi.  2,  figs.  3,  4;  pi.  3,  fig.  1. 

1968  Dactylioceras  ( Orthodactylites ) semicelatum  (Simpson);  Hoffmann,  p.  6,  pi.  2,  figs.  1,  2. 

1968  Dactylioceras  ( Orthodactylites ) eikenbergi  Hoffmann,  p.  8,  pi.  1,  fig.  2. 

1968  Dactylioceras  ( Orthodactylites ) wunnenbergi  Hoffmann,  p.  7,  pi.  1,  fig.  1. 

1968  Dactylioceras  ernsti  Lehmann,  p.  46,  pi.  17,  figs.  5,  6;  pi.  19,  figs.  2,  4. 

1971  Dactylioceras  ( Orthodactylites ) anguinum  (Reinecke);  Pinna  and  Levi-Setti,  p.  90,  pi.  2, 
figs.  1,2,5. 

1971  Dactylioceras  ( Orthodactylites ) semicelatum  (Simpson);  Pinna  and  Levi-Setti,  p.  90,  pi.  2, 
figs.  3,4,  15. 

1973  Dactylioceras  ( Orthodactylites ) semicelatum  (Simpson);  Howarth,  p.  262,  pi.  6,  fig.  1;  pi.  7, 
figs.  1,  2;  pi.  8,  figs.  1-4;  pi.  9,  figs.  1-3. 

Occurrence.  Dorset  coast:  Marlstone  Rock  Bed  layer  P,  fairly  frequent;  Somerset:  bed  2 (Hamlet  1922)  at 
Barrington,  Ilminster,  about  four  specimens  known;  north  Oxfordshire  and  Northamptonshire:  abundant  in  the 
top  of  the  Marlstone  Rock  Bed  at  many  localities  from  south  and  east  of  Banbury  to  Byfield,  Daventry,  and 
Northampton;  Leicestershire:  abundant  in  the  top  of  the  Marlstone  Rock  Bed  in  the  Tilton  area  and  common  at 
Harston;  Lincoln:  shales  of  beds  12  and  14  (Howarth  1958,  p.  xi);  north  Lincolnshire  and  south  Humberside: 
shales  2-3-4  0 m above  the  Marlstone  Rock  Bed  at  Kirton  in  Lindsey  and  Roxby. 

Discussion.  The  occurrence  of  Dactylioceras  ( Orthodactylites ) semicelatum  in  the  Grey  Shales 
Formation  of  the  Yorkshire  coast  has  already  been  described  in  detail  by  Howarth  (1973,  p.  262),  and 


EXPLANATION  OF  PLATE  80 

Figs.  1-12.  Dactylioceras  (Orthodactylites)  semicelatum  ( Simpson).  All  from  top  0-1  m of  the  Marlstone  Rock 
Bed  (‘Transition  Bed’),  Semicelatum  Subzone,  Tenuicostatum  Zone,  of  the  Banbury  area,  Northamptonshire. 
1 , 2,  5,  6,  King’s  Sutton,  6 km  SE  of  Banbury,  BM  C. 67697,  C.67376.  3,  4,  Middleton  Cheney,  4 km  ENE  of 
Banbury,  originally  figured  Buckman  (1926a,  pi.  654)  as  holotype  of  Orthodactylites  directus,  IGS  GSM 
47847.  7,  8,  Adderbury,  6 km  SSE  of  Banbury,  IGS  GSM  22566.  9,  10,  Chipping  Warden,  10  km  NE  of 
Banbury,  BM  C. 67388.  11,  12,  Copredy,  6 km  north  of  Banbury,  originally  figured  J.  Sowerby  (1819,  p.  41, 
pi.  222,  fig.  1),  paralectotype  of  Ammonites  annulatus , BM  C.40125.  All  figures  x 1. 


PLATE  80 


howarth,  ammonite  Dactylioceras 


648 


PALAEONTOLOGY,  VOLUME  23 


reference  should  be  made  to  that  paper  for  an  account  of  the  type  specimen,  the  diagnosis  and  the 
general  description  of  the  species.  Outside  Yorkshire,  the  commonest  occurrence  is  in  the  topmost 
part  of  the  Marlstone  Rock  Bed  (the  ‘Transition  Bed’)  in  Northamptonshire  and  Leicestershire.  The 
name  Orthodactylites  directus  Buckman  (1926<z,  pi.  654)  has  always  been  used  for  these  examples 
previously  (including  Howarth  1973,  pp.  266-267).  However,  analysis  of  the  west  Northamptonshire 
fauna  shows  that  it  agrees  closely  with  the  Yorkshire  fauna  of  D.  ( O .)  semicelatum  in  all  characters. 
Whorl  dimensions  and  rib-density  are  expressed  graphically  in  text-figs.  2 and  3,  where  it  can  be  seen 
that  there  are  no  significant  differences  from  the  Yorkshire  fauna,  and  that  the  Yorkshire  holotype 
occupies  an  approximately  central  position  within  the  variation  of  the  Northamptonshire  fauna.  An 
average  specimen  from  the  Marlstone  Rock  Bed  of  Northamptonshire  is  figured  in  Plate  80,  figs.  1,2, 
an  example  with  higher  whorls  and  more  rectiradiate  ribs  in  Plate  80,  figs.  5,  6,  and  a more  involute 
example  with  higher  whorls  in  Plate  80,  figs.  9,  10.  Text-figs  2 and  3 also  show  that  the  holotype  of  D. 
directum  (PI.  80,  figs.  3,  4)  is  an  extreme  form  being  more  evolute,  more  compressed,  and  more  finely 
ribbed  than  most  Northamptonshire  specimens.  Nevertheless,  it  does  fall  within  the  range  of 
variation  of  the  population,  and  it  matches  some  Yorkshire  specimens  closely  (e.g.  Howarth  1973, 
pi.  8,  fig.  1),  so  the  specific  name  directum  should  be  placed  in  synonymy  with  semicelatum. 

The  only  west  Northamptonshire  specimen  that  is  more  finely  ribbed  is  one  of  the  paralectotypes  of 
Ammonites  annulatus  J.  Sowerby  (1819,  pi.  222,  fig.  1).  Previously  (Howarth  1973,  p.  262)  it  was 
determined  as  D.  ( O .)  tenuicostatum,  but,  although  it  is  evolute  and  finely  ribbed  (PI.  80,  figs.  11,  12), 
it  has  the  characteristic  compressed  oval  (not  near-circular)  whorls  and  prorsiradiate  ribs  of  D.  ( O .) 


text-fig.  2.  Scatter  diagrams  of  whorl  dimensions 
(whorl  height,  whorl  breadth,  and  umbilical  width, 
plotted  against  diameter)  for  fifty-nine  specimens  of 
Dactylioceras  ( Orthodactylites ) semicelatum  (Simp- 
son) from  the  top  of  the  Marlstone  Rock  Bed  in  north 
Oxfordshire,  Northamptonshire,  and  Leicestershire. 
The  dashed  lines  are  the  envelopes  of  these  points, 
while  the  solid  lines  are  the  envelopes  of  the  scatter 
diagrams  of  the  Yorkshire  coast  population  of  the 
same  species  (from  Howarth  1973,  p.  259,  fig.  5). 


HOWARTH:  AGE  OF  MARLSTONE  ROCK  BED 


649 


semicelatum  and  is  matched  closely  by  several  specimens  from  bed  30  in  Yorkshire  (e.g.  Howarth 
1973,  pi.  8,  figs.  1,  2,  and  BM  C. 77304).  Another  west  Northamptonshire  specimen  was  made  the 
holotype  of  O.  mitis  Buckman  (1927a,  pi.  738):  it  also  is  not  typical  of  the  Northamptonshire  fauna, 
being  more  evolute  than  most  specimens  and  it  has  flat  whorl  sides  and  widely  spaced  ribs  near  the 
aperture  (text-figs.  2,  3;  PI.  8 1 , figs.  7-9).  It  is  an  incomplete  immature  specimen  44  mm  diameter,  and 
it  is  matched  very  closely  by  two  specimens  from  bed  28  in  Yorkshire  and  by  some  from  Harston, 
Leicestershire  (e.g.  PI.  82,  figs.  11,  12).  These  are  only  another  form  in  the  variation  of  the  species, 
with  a different  combination  of  characters,  being  evolute  with  fewer  ribs,  and  O.  mitis  should  also  be 
placed  in  synonymy  with  D.  ( O .)  semicelatum.  A larger  west  Northamptonshire  example  with  similar 
widely  spaced  ribs  is  figured  in  Plate  80,  figs.  7,  8.  It  is  one  of  the  few  complete  adults  that  are  known 
from  the  Marlstone  Rock  Bed,  and  has  a mouth  border  at  55  mm  diameter.  A specimen  from  Tilton, 
Leicestershire  (C. 80278),  has  a mouth  border  at  54  mm  diameter,  and  two  other  Northamptonshire 
and  Harston  specimens  are  97  and  99  mm  diameter  at  their  adult  mouth  borders  respectively.  This 
54-99  mm  range  compares  with  an  adult  diameter  range  of  75- 1 20  mm  for  the  Y orkshire  coast  fauna. 
A typical  example  from  the  top  0T  m of  the  Marlstone  Rock  Bed  at  Tilton  is  figured  in  Plate  81, 
figs.  5,  6.  Two  small  and  indifferently  preserved  Dorset  coast  specimens  were  figured  previously 
(Howarth  1957,  p.  197,  pi.  17,  figs.  5,  6);  a large,  better  preserved  example  is  figured  in  Plate  81, 
figs.  3, 4,  which  is  a typical  involute  specimen  with  the  high,  oval  whorls  of  the  species.  At  Barrington, 
Somerset,  specimens  occur  in  a bed  about  0-2  m above  the  Marlstone  Rock  Bed,  and  the  best  one  is 
figured  in  Plate  81,  figs.  1,  2. 

The  only  occurrence  of  D.  ( O .)  semicelatum  outside  Britain  that  was  not  dealt  with  in  the  Y orkshire 
coast  paper  (Howarth  1973)  consists  of  those  specimens  in  north-west  Germany  described  as  D.  ernsti 
by  Lehmann  (1968,  p.  46,  pi.  17,  figs.  5, 6;  pi.  19,  figs.  2, 4;  also  figured  by  Hoffmann  1968)  and  smaller 
specimens  figured  by  Hoffmann  and  Martin  (1960).  These  show  all  the  usual  characters  of  D.  (O.) 
semicelatum,  and  the  holotype  of  D.  ernsti  has  whorl  proportions  and  rib-density  that  are  close  to  the 
average  of  the  Yorkshire  and  Northamptonshire  populations.  All  the  north-west  German  specimens 
come  from  the  Semicelatum  Subzone,  and  D.  ernsti  is  considered  to  be  a synonym  of  D.  ( O .) 
semicelatum. 


text-fig.  3.  Scatter  diagram  of  number  of  ribs  per 
whorl  for  seventy-one  specimens  of  Dactylioceras 
0 Orthodactylites ) semicelatum  (Simpson)  from  the  top 
of  the  Marlstone  Rock  Bed  in  north  Oxfordshire, 
Northamptonshire,  and  Leicestershire.  The  dashed 
line  is  the  envelope  of  these  points;  the  solid  line  is  the 
envelope  of  the  scatter  diagram  of  the  Yorkshire  coast 
population  of  the  same  species  (from  Howarth  1973, 
p.  261,  fig.  6). 


650 


PALAEONTOLOGY,  VOLUME  23 


Dactylioceras  ( Orthodactylites ) tenuicostatum  (Young  and  Bird) 

Plate  82,  figs.  1-10,  13,  14 

1822  Ammonites  tenuicostatus  Young  and  Bird,  p.  247,  pi.  12,  fig.  8. 

1828  Ammonites  annulatus  Sowerby;  Young  and  Bird,  p.  253,  pi.  12,  fig.  11. 

1884  Stephanoceras  annulatum  (J.  Sowerby);  Wright,  p.  475,  pi.  84,  figs.  7,  8. 

1920a  Dactylioceras  tenuicostatum  (Young  and  Bird);  Buckman,  pi.  157. 

1927a  Tenuidactylites  tenuicostatus  (Young  and  Bird);  Buckman,  pi.  157A. 

1933  Dactylioceras  tenuicostatum  (Young  and  Bird);  Arkell,  pi.  32,  fig.  6. 

1956  Dactylioceras  tenuicostatum  (Young  and  Bird);  Arkell,  pi.  33,  fig.  6. 

1957  Dactylioceras  tenuicostatum  (Young  and  Bird);  Maubeuge,  p.  208,  figs.  ?41,  42,  43. 

1961  Dactylioceras  tenuicostatum  (Young  and  Bird);  Dean,  Donovan,  and  Howarth,  pi.  72,  fig.  1. 

1973  Dactylioceras  ( Orthodactylites ) tenuicostatum  (Young  and  Bird);  Howarth,  p.  258,  pi.  5, 
figs.  1,2;  pi.  6,  figs.  2,3. 

Occurrence.  Dorset  coast:  Marlstone  Rock  Bed  layer  P,  uncommon;  Somerset:  bed  1 (Hamlet  1922)  at 
Barrington,  Ilminster,  poorly  preserved  crushed  specimens;  Northamptonshire:  Rothersthorpe,  one  specimen; 
Leicestershire:  008-0T3  m below  the  top  of  the  Marlstone  Rock  Bed  at  Harston,  abundant;  north  Lincolnshire 
and  south  Humberside:  hard  mudstone  and  1 m of  shales  above  the  Marlstone  Rock  Bed  at  Kirton  in  Lindsey 
and  Roxby,  common. 

Discussion.  A full  account  of  the  type  specimen,  diagnosis,  and  the  Yorkshire  coast  fauna  is  found  in 
Howarth  (1973,  pp.  258-262).  Outside  Yorkshire,  D.  tenuicostatum  is  much  less  widely  distributed 
than  D.  semicelatum,  and  the  only  substantial  collection  from  the  Marlstone  Rock  Bed  was  that 
obtained  from  Harston,  Leicestershire.  About  seventy  specimens  were  collected,  all  of  them 
immature  and  less  than  60  mm  maximum  diameter.  Most  have  part  of  their  body  chambers  preserved 
but  they  are  incomplete,  and  no  adult  specimens,  indicated  by  constricted  mouth  borders  or 
approximated  final  suture-lines,  were  found.  All  have  the  typical  rounded  whorl  section  and  fine  ribs 
of  D.  tenuicostatum.  An  immature  of  average  size  is  figured  in  Plate  82,  figs.  9, 10,  and  the  largest  of  58 
mm  diameter  in  Plate  82,  figs.  1 , 2.  The  top  part  of  the  Marlstone  Rock  Bed  at  Harston  is  highly 
condensed,  and  although  the  main  occurrence  of  D.  semicelatum  is  higher  up,  a few  specimens  of  the 
latter  species  are  found  at  the  same  level  as  the  highest  D.  tenuicostatum.  Specimens  of  D.  semicelatum 
are  always  separable  by  their  higher  whorls,  oval  whorl  section,  more  widely  spaced  ribs,  and  by  the 
considerably  thicker  whorls  in  some  individuals. 

A single  well-preserved  specimen  has  already  been  referred  to  (p.  641)  from  the  top  of  the 
Marlstone  Rock  Bed  at  Rothersthorpe,  5 km  south-west  of  Northampton  (PI.  82,  figs.  3,  4).  It  is 
immature,  46  mm  diameter,  and  has  a body  chamber  one  whorl  long.  About  ten  examples  of  D. 
tenuicostatum  are  known  from  layer  P of  the  Marlstone  Rock  Bed  on  the  Dorset  coast.  Again  they  are 
all  immature  or  inner  whorls  of  less  than  about  60  mm  diameter,  and  two  of  the  best  specimens  are 
figured  in  Plate  82,  figs.  5-8.  In  north  Lincolnshire  the  species  occurs  in  the  hard  mudstone  that 
overlies  the  Marlstone  Rock  Bed,  and  one  of  the  more  complete,  though  small,  specimens  is  figured  in 
Plate  82,  figs.  13,  14. 


EXPLANATION  OF  PLATE  81 

Figs.  1-11.  Dactylioceras  ( Orthodactylites ) semicelatum  (Simpson).  1,  2,  bed  2 (Hamlet  1922),  0-2  m above 
Marlstone  Rock  Bed,  Barrington  Quarry,  near  Ilminster,  Somerset,  IGS  GSM  31612.  3,  4,  Marlstone  Rock 
Bed  layer  P,  Seatown,  Dorset,  BM  C.  17548.  5,  6,  Marlstone  Rock  Bed,  top  01  m,  Tilton  Railway  Cutting, 
Leicestershire,  BM  C.  80277.  7-9,  top  of  Marlstone  Rock  Bed  (‘Transition  Bed’),  Byfield,  Northamptonshire, 
originally  figured  Buckman  (1927a,  pi.  738)  as  holotype  of  Orthodactylites  mitis,  IGS  GSM  38384.  10,  11, 
Marlstone  Rock  Bed,  0 08  m below  top,  Harston  Quarry,  north  Leicestershire,  BM  C.80169.  All  figures  x 1. 


PLATE  81 


howarth,  ammonite  Dactylioceras 


652 


PALAEONTOLOGY,  VOLUME  23 


Dactylioceras  ( Orthodactylites ) clevelandicum  Howarth 
Plate  82,  figs.  15,  16 

1973  Dactylioceras  ( Orthodactylites ) clevelandicum  Howarth,  pp.  257-258,  pi.  3,  figs.  1-3;  pi.  4, 
figs.  1,2;  pi.  5,  fig.  3. 

Occurrence.  About  eight  specimens  known  in  bed  21  at  Kirton  in  Lindsey,  north  Lincolnshire,  and  an  equivalent 
horizon  in  the  near-by  Worlaby  borehole. 

Discussion.  The  most  difficult  problem  in  describing  the  Tenuicostatum  Zone  Dactylioceratidae  that 
occur  in  the  Marlstone  Rock  Bed  area  in  England  is  the  identification  of  the  well-preserved  specimens 
in  the  hard  mudstone  and  the  calcareous  nodules  (bed  21)  that  overlie  the  Rock  Bed  at  Kirton  in 
Lindsey,  north  of  Lincoln.  Specimens,  though  well  preserved,  are  not  very  numerous,  and  considered 
on  their  own  they  could  be  a condensed  mixture  of  D.  semicelatum,  D.  tenuicostatum , and  D.  cleve- 
landicum. Some  limit  to  the  age  range  can  be  obtained,  however,  from  the  ammonites  in  the  overlying 
beds,  for  crushed  Dactylioceras  that  appear  to  be  D.  tenuicostatum  occur  in  the  overlying  1-3  m of 
shale,  then  D.  semicelatum  appears  in  the  next  higher  1 -7  m of  shale,  and  unmistakable  specimens  of 
Tiltoniceras  occur  in  the  next  3-5  m.  These  ammonites  are  in  the  correct  stratigraphical  sequence  for 
the  Tenuicostatum  and  Semicelatum  Subzones.  So  it  is  probable  that  only  the  lower  part  of  the 
Tenuicostatum  Subzone,  together  with  lower  horizons,  occurs  in  the  hard  mudstone.  This  mudstone, 
and  especially  the  calcareous  nodules  within  it,  contains  a number  of  small  or  fragmentary  specimens 
of  D.  tenuicostatum  (PI.  82,  figs.  13,  14),  and  also  a small  collection  of  larger  and  better-preserved 
individuals  that  are  identified  with  D.  clevelandicum.  One  of  the  best  specimens  from  the  mudstone  at 
Kirton  in  Lindsey  is  figured  in  Plate  82,  figs.  15,  16.  Although  it  is  like  D.  semicelatum  in  some 
respects,  it  has  rectiradiate  ribs,  not  the  prorsiradiate  ribs  of  compressed  specimens  of  D. 
semicelatum.  Nor  does  it  have  the  typical  oval  whorl  section  of  the  latter  species.  Other  examples  in 
this  bed  have  a range  of  variation  from  rounded  whorls  with  fine  ribs,  to  depressed  whorls  with  coarse 
ribs  and  tubercles.  The  few  that  are  measurable  all  fall  within  the  ranges  for  whorl  dimensions  and 
rib-density  measured  for  the  Yorkshire  coast  population  of  D.  clevelandicum  (Howarth  1973, 
pp.  259,  261).  No  examples  of  this  species  have  been  found  anywhere  else  in  the  Marlstone  Rock  Bed 
area. 


EXPLANATION  OF  PLATE  82 

Figs.  1-10,  13,  14.  Dactylioceras  ( Orthodactylites ) tenuicostatum  (Young  and  Bird).  1,  2,  9,  10,  Marlstone 
Rock  Bed,  0T  m below  top,  Harston  Quarry,  north  Leicestershire,  BM  C.80 1 22, 80 100.  3, 4,  Marlstone  Rock 
Bed,  immediately  below  ‘Transition  Bed’,  Rothersthorpe,  5 km  SW  of  Northampton,  BM  C.82051.  5-8, 
Marlstone  Rock  Bed  layer  P,  Doghouse  Cliff,  Seatown,  Dorset,  NMW  26.135  G124and  G5.2.  13, 14,  bed  21 
(Howarth  and  Rawson  1965),  0-3  m above  Marlstone  Rock  Bed,  quarry  2 km  north  of  Kirton  in  Lindsey, 
north  Lincolnshire,  BM  C. 73560.  All  figures  x 1. 

Figs.  11,  12.  Dactylioceras  ( Orthodactylites ) semicelatum  (Simpson).  Marlstone  Rock  Bed,  0-08  m below  top, 
Harston  quarry,  north  Leicestershire,  BM  C.80 173,  x 1. 

Figs.  15,  16.  Dactylioceras  ( Orthodactylites ) clevelandicum  Howarth.  Bed  21  (Howarth  and  Rawson  1965), 
0-3  m above  Marlstone  Rock  Bed,  quarry  2 km  north  of  Kirton  in  Lindsey,  north  Lincolnshire,  BM 
C. 73561,  x 1. 

Figs.  17,  18.  Dactylioceras  ( Orthodactylites ) crosbeyi  (Simpson).  Marlstone  Rock  Bed,  0-23  m below  top, 
Harston  quarry,  north  Leicestershire,  BM  C.80 170,  x 1. 


PLATE  82 


'/////% 


howarth,  ammonite  Dactylioceras 


654 


PALAEONTOLOGY,  VOLUME  23 


Dactylioceras  ( Orthodactylites ) crosbeyi  (Simpson) 

Plate  82,  figs.  17,  18 

1843  Ammonites  crosbeyi  Simpson,  p.  22. 

1855  Ammonites  crosbeyi  Simpson,  p.  58. 

1884  Ammonites  crosbeyi  Simpson,  p.  90. 

1912a  Coeloceras  crosbeyi  (Simpson);  Buckman,  pi.  60. 

71957  Dactylioceras  pseudosemicelatum  Maubeuge,  p.  193,  pi.  3,  fig.  6. 

71957  Dactylioceras podagrosum  Maubeuge,  p.  193,  pi.  4,  fig.  7. 

1973  Dactylioceras  ( Orthodactylites ) crosbeyi  (Simpson);  Howarth,  p.  255,  pi.  1,  figs.  2-4;  pi.  2, 
figs.  1-4. 

Occurrence.  North  Leicestershire:  0-23  m below  the  top  of  the  Marlstone  Rock  Bed,  Harston  quarry,  one 
specimen. 

Discussion.  This  broken  half  ammonite  is  about  74  mm  diameter,  and  the  final  one-third  of  a whorl  is 
probably  body-chamber.  It  has  relatively  high  and  broad  whorls  that  are  about  one-quarter  involute, 
and  the  whorl  section  has  an  evenly  rounded  venter.  The  preservation  is  mainly  as  an  internal  cast,  so 
the  ribbing  is  of  very  low  relief,  and  consists  of  prorsiradiate  primary  ribs,  about  half  of  which 
bifurcate  at  the  ventro-lateral  edge.  The  ribs  on  the  venter  swing  slightly  more  forwards,  and  only  the 
slightest  traces  of  ventro-lateral  tubercles  are  present.  At  74  mm  diameter  the  whorl  height  is  21  -5  mm 
and  the  breadth  is  2 TO  mm,  and  these  whorl  dimensions  agree  well  with  those  of  Yorkshire  coast 
specimens  of  D.  ( O .)  crosbeyi.  It  compares  well  with  the  more  compressed,  more  finely  ribbed 
examples  of  the  species  such  as  were  figured  by  Howarth  (1973,  pi.  1,  fig.  2;  pi.  2,  fig.  2).  The  whorl 
height  and  the  amount  of  overlap  of  the  whorls  are  both  too  large  for  D.  ( O .)  clevelandicum.  The 
specimen  occurs  0- 1 m below  a rich  population  of  D.  ( O .)  tenuicostatum  in  the  Marlstone  Rock  Bed  at 
Harston,  a stratigraphical  position  that  agrees  with  its  occurrence  in  Yorkshire.  No  trace  was  found 
at  Harston  of  the  intervening  species  D.  (O.)  clevelandicum.  No  other  examples  of  D.  (O.)  crosbeyi 
have  been  found  outside  Yorkshire. 


CONCLUSIONS 

The  ‘Transition  Bed’  is  the  weathered  or  altered  top  of  the  Marlstone  Rock  Bed.  The  main  change  is 
oxidation  of  the  green  ferrous  minerals  to  limonite,  and  associated  partial  decalcification  leaves  the 
bed  crumbly  or  ‘sandy’  in  some  places.  The  weathering  occurred  partly  before  deposition  of  the  over- 
lying  beds  in  some  areas,  e.g.  Banbury  and  west  Northamptonshire,  though  at  Tilton  most  of  the 
weathering  is  more  recent.  Another  type  of  alteration  that  took  place  before  deposition  of  overlying 
beds,  was  the  pyritization  of  the  bed  in  the  Harston  area,  Leicestershire.  There  is  no  evidence  that  the 
bed  is  otherwise  mineralogically  different  from  the  Marlstone  Rock  Bed,  and  there  is  no  sedimentary 
discontinuity  at  its  base.  The  term  Marlstone  Rock  Bed  should  be  applied  to  the  whole  of  the  bed. 

In  south  Dorset  and  from  north  Oxfordshire  to  south  Lincolnshire  the  Marlstone  Rock  Bed  was 
deposited  during  all  the  period  represented  by  the  Spinatum  and  Tenuicostatum  Zones,  and  there  is 
no  lithological  division  between  the  two  zones.  The  ammonite  faunas  at  the  boundary  are  poor,  but 
generally  the  top  1-3  m belongs  to  the  Tenuicostatum  Zone  and  the  bottom  3-6  m to  the  Spinatum 
Zone.  From  north  Somerset  to  south  Oxfordshire  there  is  no  ammonite  evidence  for  the  age  of  the  top 
of  the  bed. 

The  following  ammonite  faunas  have  been  found  in  the  Marlstone  Rock  Bed: 

(a)  Dactylioceras  semicelatum  (D.  directum  is  a synonym)  and  Tiltoniceras  antiquum  of  the 
Semicelatum  Subzone.  This  is  abundant  at  many  localities  and  is  the  fauna  of  the  ‘Transition  Bed’. 

( b ) D.  tenuicostatum  of  the  Tenuicostatum  Subzone.  Abundant  only  at  Harston,  Leicestershire, 
present  on  the  Dorset  coast,  and  rare  elsewhere. 

(c)  D.  crosbeyi  of  the  Clevelandicum  Subzone.  One  specimen  at  Harston. 

(d)  Protogrammoceras  paltum  of  the  Paltum  Subzone.  Only  on  the  Dorset  coast. 


HOWARTH:  AGE  OF  MARLSTONE  ROCK  BED 


655 


(e)  Pleuroceras  spp.  of  the  Spinatum  Zone.  Abundant  in  Dorset,  Somerset,  and  Gloucestershire. 
Much  rarer  from  Oxfordshire  to  north  Humberside,  but  sufficient  are  known  to  show  that  both  the 
Apyrenum  and  Hawskerense  Subzones  are  present. 

From  north  of  Lincoln  to  north  Humberside  deposition  of  the  Marlstone  Rock  Bed  stopped  at  the 
end  of  the  Spinatum  Zone,  and  ammonites  of  all  four  subzones  of  the  Tenuicostatum  Zone  occur  in 
an  overlying,  lithologically  distinct,  hard  mudstone  and  in  shales  above.  The  latter  are  similar  to  the 
Grey  Shales  Formation  of  the  Yorkshire  coast. 

The  change  from  the  lower,  regressive  ironstone/limestone  facies  to  the  upper  transgressive 
clays/shales-with-nodules  facies  (Hallam  1967,  pp.  431-440)  did  not  occur  simultaneously  in  Britain. 
In  Yorkshire  it  occurred  at  the  top  of  the  Spinatum  Zone  at  the  Upper  Pliensbachian/Toarcian  (i.e. 
Middle/Upper  Lias)  boundary,  but  from  Dorset  to  south  of  Lincoln  it  occurred  at  the  top  of  the 
Tenuicostatum  Zone.  In  a small  transitionary  area  between  north  of  Lincoln  and  Market  Weighton 
the  change  took  place  in  the  middle  of  the  Tenuicostatum  Subzone.  The  extent  of  the  time  disparity 
for  the  facies  change  may  be  judged  from  the  fact  that  14  m of  Grey  Shales  Formation  on  the  York- 
shire coast  were  being  deposited  while  1 -3  m of  Marlstone  Rock  Bed  was  being  deposited  in  England 
south  of  Lincoln. 

Acknowledgements.  The  author  wishes  to  thank  the  British  Steel  Corporation  who  originally  allowed  access  to 
the  Harston,  Denton  Park,  and  Roxby  Quarries.  The  paper  has  benefited  from  discussions  with  Professor 
A.  Hallam,  and  with  Mr.  M.  D.  Jones  of  Leicester  City  Museum  regarding  the  Tilton  and  Harston  exposures.  As 
well  as  the  collections  at  the  British  Museum  (Natural  History)  (specimens  with  prefix  BM),  specimens  were  also 
examined  at  the  Institute  of  Geological  Sciences,  London  (prefix  IGS),  through  the  kindness  of  Dr.  H.  C.  Ivimey- 
Cook,  and  borrowed  from  the  National  Museum  of  Wales,  Cardiff  (prefix  NMW)  through  Dr.  D.  A.  Bassett. 


REFERENCES 

ager,  D.  v.  1956a.  Some  new  Liassic  Terebratuloids.  Proc.  Geol.  Ass.  67,  1-14,  pi.  1. 

— 19566.  A monograph  of  the  British  Liassic  Rhynchonellidae.  Part  1,  i-xxvi,  1-50,  pis.  1-4.  Palaeontogr. 
Soc.  [Monogr.]. 

arkell,  w.  j.  1933.  The  Jurassic  System  in  Great  Britain.  Oxford,  xii  + 681  pp.,  41  pis. 

— 1947.  The  Geology  of  Oxford.  Oxford.  267  pp. 

— 1956.  Jurassic  Geology  of  the  World.  Edinburgh  and  London,  xv  + 806  pp.,  46  pis. 

buckman,  s.  s.  1909a- 1930a.  Yorkshire  Type  Ammonites , 1,  2,  and  Type  Ammonites , 3-7.  London.  790  pis. 

— 19106.  Certain  Jurassic  (Lias-Oolite)  strata  of  south  Dorset;  and  their  correlation.  Q.  Jl  geol.  Soc.  Lond.  66, 
52-89. 

— 19226.  Jurassic  Chronology;  II.  Preliminary  studies.  Certain  Jurassic  strata  near  Eypesmouth  (Dorset); 
the  Junction  Bed  of  Watton  Cliff  and  associated  rocks.  Ibid.  78,  378-475. 

dean,  w.  T.,  donovan,  d.  t.  and  howarth,  M.  K.  1961.  The  Liassic  ammonite  zones  and  subzones  of  the  north- 
west European  Province.  Bull.  Br.  Mus.  nat.  Hist.,  Geol.  4,  435-505,  pis.  63-75. 
fox-strangways,  c.  1903.  The  Geology  of  the  country  near  Leicester.  Mem.  geol.  Surv.  U.K.  122  pp.,  2 pis. 
hallam,  a.  1955.  The  palaeontology  and  stratigraphy  of  the  Marlstone  Rock-bed  in  Leicestershire.  Trans. 
Leicester  lit.  phil.  Soc.  49,  17-35. 

— 1967.  An  environmental  study  of  the  Upper  Domerian  and  Lower  Toarcian  in  Great  Britain.  Phil.  Trans. 
R.  Soc.  (B)  252,  393-445,  pi.  20. 

— 1968.  The  Lias.  In  sylvester-bradley,  p.  c.  and  ford,  t.  d.  (eds.).  The  Geology  of  the  East  Midlands. 
Leicester.  201-211. 

— 1972.  Diversity  and  density  characteristics  of  Pliensbachian-Toarcian  molluscan  and  brachiopod  faunas 
of  the  north  Atlantic  margins.  Lethaia,  5,  389-412. 

hamlet,  j.  1922.  On  sections  in  the  Lias  exposed  in  two  quarries  at  Barrington.  Proc.  Somerset,  archaeol.  nat. 
Hist.  Soc.  67,  72-75. 

hoffmann,  k.  1968.  Neue  Ammonitenfunde  aus  dem  tieferen  Unter-Toarcium  (Lias  e)  des  nordlichen 
Harzvorlandes  und  ihre  feinstratigraphische  Bedeutung.  Geol.  Jb.  85,  1-32,  pis.  1-5. 

— and  martin,  p.  R.  1960.  Die  Zone  des  Dactylioceras  tenuicostatum  (Toarcien,  Lias)  in  NW-  und 
SW-Deutschland.  Palaeont.  Z.  34,  103-149,  pis.  8-12. 

howarth,  m.  k.  1957.  The  Middle  Lias  of  the  Dorset  Coast.  Q.  J!  geol.  Soc.  Lond.  113,  185-204,  pi.  17. 


656 


PALAEONTOLOGY,  VOLUME  23 


howarth,  m.  k.  1958.  A monograph  of  the  ammonites  of  the  Liassic  family  Amaltheidae  in  Britain.  Part  1,  i-xvi, 
1-26,  pis.  1-4.  Palaeontogr.  Soc.  [ Monogr .]. 

— 1973.  The  stratigraphy  and  ammonite  fauna  of  the  Upper  Liassic  Grey  Shales  of  the  Yorkshire  coast. 
Bull.  Br.  Mus.  nat.  Hist.,  Geol.  24,  235-277,  pis.  1-9. 

— 1978.  The  stratigraphy  and  ammonite  fauna  of  the  Upper  Lias  of  Northamptonshire.  Ibid.  29,  235-288, 
pis.  1-9. 

— and  rawson,  p.  F.  1965.  The  Liassic  succession  in  a clay  pit  at  Kirton  in  Lindsey,  north  Lincolnshire.  Geol. 
Mag.  102,  261-266. 

hull,  E.  1857.  The  geology  of  the  country  around  Cheltenham.  Mem.  geol.  Surv.  U.K.  104  pp.,  2 pis. 
jackson,  J.  F.  1922.  Sections  of  the  Junction  Bed  and  contiguous  deposits.  Q.  Jl  geol.  Soc.  Lond.  78,  436-448. 

— 1926.  The  Junction-Bed  of  the  Middle  and  Upper  Lias  on  the  Dorset  coast.  Ibid.  82,  490-525,  pis.  33,  34. 
lehmann,  u.  1968.  Strati graphie  und  Ammonitenfiihrung  der  Ahrensburger  Glazial-Geschiebe  aus  dem  Lias 

epsilon  (=  Unt.  Toarcium).  Mitt.  geol.  Stlnst.  Hamb.  37,  41-68,  pis.  17-20. 
maubeuge,  p.  L.  1957.  Les  ammonites  de  la  zone  a Dactylioceras  semicelatum-tenuicostatum  dans  l’Est  de  la 
France  et  plus  specialement  dans  le  Grande-Duche  de  Luxembourg.  Archs  Inst,  gr.-duc.  Luxemb.  (n.s.),  24, 
189-226,  pis.  1-30. 

penny,  l.  f.  and  rawson,  p.  f.  1969.  Field  meeting  in  east  Yorkshire  and  north  Lincolnshire.  Proc.  Geol.  Ass.  80, 
193-218. 

pinna,  g.  and  levi-setti,  f.  1971.  I Dactylioceratidae  della  Provincia  Mediterranea  (Cephalopoda, 
Ammonoidea).  Memorie  Soc.  ital.  Sci.  nat.  19,  47-136,  pis.  1-12. 
pringle,  J.  and  templeman,  a.  1922.  Two  new  sections  in  the  Middle  and  Upper  Lias  at  Barrington,  near 
Ilminster,  Somerset.  Q.  Jl  geol.  Soc.  Lond.  78,  450-451. 

Richardson,  G.  1979.  The  Mesozoic  stratigraphy  of  two  boreholes  near  Worlaby,  South  Humberside.  Bull, 
geol.  Surv.  Gt  Br.  58. 

Richardson,  l.  1906.  On  a section  of  Middle  and  Upper  Lias  rocks  near  Evercreech,  Somerset.  Geol.  Mag.  (5) 
3, 368-369. 

— 1909.  On  some  Middle  and  Upper  Lias  sections  near  Batcombe,  Somerset.  Ibid.  (5)  6,  540-542. 

— 1929.  The  country  around  Moreton  in  Marsh.  Mem.  Geol.  Surv.  U.K.  162  pp. 
simpson,  M.  1843.  A Monograph  of  the  Ammonites  of  the  Yorkshire  Lias.  London.  60  pp. 

— 1855.  The  Fossils  of  the  Yorkshire  Lias;  Described  from  Nature.  London  and  Whitby.  149  pp. 

1884.  Ibid.  London  and  Whitby.  2nd  edn.,  xxiii  + 256  pp. 

sowerby,  j.  1819.  The  Mineral  Conchology  of  Great  Britain.  London.  Vol.  3,  pis.  222-253. 

spath,  l.  f.  1922.  Upper  Lias  succession  near  Ilminster,  Somerset.  Q.  Jl  geol.  Soc.  Lond.  78,  449-450. 

— 1942.  The  ammonite  zones  of  the  Lias.  Geol.  Mag.  79,  264-268. 

— 1956.  The  Liassic  ammonite  faunas  of  the  Stowell  Park  Borehole.  Bull.  geol.  Surv.  Gt  Br.  11,  140-164, 
pis.  9,  10. 

Thompson,  b.  1889.  The  Middle  Lias  of  Northamptonshire.  London.  150  pp.  (Reprinted  from  Midi.  Nat.  8 
(1885)— 12  (1889)). 

— 1892.  Report  of  the  Committee  ...  to  work  on  the  very  fossiliferous  Transition  Bed  between  the  Middle 
and  Upper  Lias  in  Northamptonshire.  Rep.  Br.  zlsi.  Advmt  Sci.  for  1890  (Cardiff  1891),  334-351  (Reprinted 
J.  Northampt.  nat.  Hist.  Soc.  7,  35-57  (1892)). 

trueman,  a.  E.  1918.  The  Lias  of  south  Lincolnshire.  Geol.  Mag.  (5)  5,  64-73,  101-111. 
walford,  a.  e.  1878.  On  some  Middle  and  Upper  Lias  beds,  in  the  neighbourhood  of  Banbury.  Proc.  Warwick. 
Nat.  Archaeol.  Fid  Club,  suppl.  for  1878,  1-23. 

— 1899.  The  Lias  Ironstone  of  North  Oxfordshire  ( around  Banbury).  London  and  Banbury.  36  pp. 
whitehead,  t.  h.  et  al.  1952.  The  Liassic  Ironstones.  Mem.  geol.  Surv.  U.K.  21 1 pp.,  8 pis. 

wilson,  e.  and  crick,  w.  d.  1889.  The  Liassic  Marlstone  of  Tilton,  Leicestershire.  Geol.  Mag.  (3)  6,  296-305, 
337-342,  pis.  9,  10. 

woodward,  H.  B.  1893.  The  Lias  of  England  and  Wales  (Y orkshire  excepted).  The  Jurassic  Rocks  of  Britain,  3. 
Mem.  geol.  Surv.  U.K.  399  pp. 

wright,  T.  1884.  A monograph  on  the  Lias  ammonites  of  the  British  Islands.  Part  7,  441-480,  pis.  78-87. 
Palaeontogr.  Soc.  [Monogr.]. 

young,  G.  M.  and  bird,  j.  1822.  A Geological  Survey  of  the  Yorkshire  Coast:  Describing  the  Strata  and  Fossils 
occurring  between  the  Humber  and  the  Tees,  from  the  German  Ocean  to  the  Plain  of  York.  Whitby.  336  pp., 
17  pis. 

1828.  Ibid.  2nd  edn.,  enlarged.  Whitby.  368  pp.,  17  pis. 


M.  k.  howarth 


Department  of  Palaeontology 

Typescript  received  20  March  1979  British  Museum  (Natural  History) 

Revised  typescript  received  10  October  1979  Cromwell  Road,  London,  SW7  5BD 


JURASSIC  ARAUCARI AN  CONE  FROM 
SOUTHERN  ENGLAND 

by  RUTH  A.  STOCKEY 


Abstract.  A well-preserved  araucarian  cone  measuring  4-5  x 5-0  cm  is  described  from  Jurassic  age  limestone 
from  near  Osmington  Mills,  Dorset.  Four  pieces  of  cone  material  representing  a single  specimen  are  somewhat 
flattened  and  lignitic,  with  intact  seed  and  cone-scale  tissues.  The  cone  axis  and  bract  apophyses  are  replaced 
with  a calcitic  matrix.  Helically  arranged  cone-scale  complexes  with  a prominent  ligular  sulcus  surround  a wide 
pith.  One  recurved  wingless  ovule  0-8  cm  long  is  deeply  sunken  into  the  cone-scale  tissue.  Seed  integuments  are 
relatively  mature  and  contain  three  distinct  layers  the  most  prominent  of  which  is  the  sclerotesta  constructed  of 
interlocking  zig-zag  sclereids.  The  nucellus,  in  some  cases  still  cellular,  is  free  from  the  integuments  except  at  the 
chalaza  and  has  the  characteristic  wavy  apex  common  to  extant  araucarians  at  a comparable  developmental 
stage.  A well-developed  vascular  system  like  that  in  Araucaria  bidwillii  Hooker  is  present  near  the  seed  chalaza. 
Cellular  megagametophyte  and  embryos  are  present  within  some  seeds.  The  specimen  is  described  as  a new 
species,  A.  brownii  sp.  nov.  in  which  the  cone  structure  most  closely  resembles  that  of  the  section  Bunya  of  the 
genus  Araucaria.  This  discovery  extends  the  range  of  this  section  to  the  Northern  Hemisphere  during  the 
Mesozoic. 

The  Araucariaceae,  an  extant  conifer  family  with  a very  restricted  distribution,  has  often  been  con- 
sidered primitive  among  conifer  families.  Two  genera,  Agathis  and  Araucaria,  grow  as  natives  in 
South  America,  Australia,  New  Caledonia,  New  Guinea,  and  a few  South  Pacific  islands.  Although 
the  group  has  only  a few  relict  species  today,  it  was  at  one  time  widespread  and  included 
numerous  species  in  the  Northern  Hemisphere  during  the  Mesozoic  Era.  Araucarian  cones  display 
what  have  been  suggested  as  primitive  characters  that  readily  distinguish  them  from  those  of  other 
conifer  groups  (Wieland  1935;  Thomson  1905a,  1907,  1913;  Eames  1913;  Wilde  and  Eames  1948, 
1952;  Burlingame  1913,  1914,  1915;  Chamberlain  1935;  Hirmer  1936;  Seward  and  Ford  1906).  The 
genus  Araucaria,  usually  believed  to  be  the  more  primitive  of  the  two  genera,  has  ovulate  cones  with 
large  bracts  and  partially  fused  ovuliferous  scales.  Agathis  exhibits  cone-scales  composed  of  com- 
pletely fused  bracts  and  scales,  believed  to  be  a derived  condition  (Eames  1913).  Well-preserved  fossil 
conifer  cones  of  any  type  are  rare;  however,  a few  well-preserved  araucarian  fossils  have  been  found, 
and  these  have  revealed  important  information  about  the  geologic  history  of  this  family,  its 
distribution,  and  reproductive  biology  (Kendall  1949;  Wieland  1935;  Darrow  1936;  Calder  1953; 
Stockey  1975,  1977,  1978;  Vishnu-Mittre  1954).  The  uniquely  preserved  fossil  conifer  cone  reported 
here  is  closely  compared  with  other  fossil  and  living  araucarians.  The  information  obtained  has 
proved  useful  in  elucidating  phylogenetic  trends  within  the  family,  and  in  particular,  within  the  genus 
Araucaria. 


MATERIALS  AND  METHODS 

The  cone  was  found  in  1973  by  Mr.  P.  A.  Brown  of  Dorset  in  a block  of  limestone  lying  on  the 
beach  under  Black  Head,  west  of  Osmington  Mills.  Specimens  are  lignitic,  in  a matrix  best  described 
as  a compacted  bio-pel-micrite  with  pelecypod  shells,  fecal  pellets,  tests  of  foraminifera,  and  corals 
which  have  all  undergone  a considerable  amount  of  diagenesis.  The  recrystallized  calcite  composing 
these  fragments  is  held  together  with  a CaC03  cement  that  has  infiltrated  many  of  the  preserved  cone 
parts  including  the  seeds.  Specimens  were  prepared  for  study  by  a modified  coal  ball  peel  technique 
(Joy,  Willis,  and  Lacey  1956)  using  76  ^m  cellulose  acetate  sheets  and  by  thin  sections  after  epoxy 

I Palaeontology,  Vol.  23,  Part  3,  1980,  pp.  657-666,  pis.  83-86.) 


658 


PALAEONTOLOGY,  VOLUME  23 


infiltration  of  the  cut  face.  Some  cone  parts  were  examined  after  gold  sputter  coating  using  an 
AMR  1000  scanning  electron  microscope  at  20  kV.  A few  whole  seeds  and  cone  fragments  were 
demineralized  in  2%  HC1  overnight  and  washed  in  distilled  water.  These  parts  were  then  embedded  in 
glycol  methacrylate  and  sectioned  with  a rotary  microtome  after  a technique  by  Robison  and  Miller 
(1975). 

Since  the  cone  was  not  found  in  place  its  exact  age  needs  some  discussion.  The  area  west  of 
Osmington  Mills,  Dorset,  to  Black  Head  where  the  cone  was  found  consists  of  cliffs  composed  of 
Corallian  and  Kimmeridge  Clay  sediments.  These  correspond  to  the  Oxfordian  and  Kimmeridgian 
Stages  respectively  of  the  Upper  Jurassic  (Arkell  1947).  The  cone  most  likely  comes  from  the 
Osmington  Oolite  Series  which  is  exposed  along  the  shore  west  to  Shortlake.  The  beds  within  this 
series  contain  several  clay  layers  with  nodules,  in  addition  to  the  oolites  and  marlstones  (Arkell  1947). 
The  matrix  surrounding  the  specimen  closely  resembles  these  beds.  Its  age  is  therefore  certainly 
Upper  Jurassic,  and  probably  Upper  Oxfordian. 


SYSTEMATIC  DESCRIPTION 

Order  coniferales 
Family  araucariaceae 
Genus  araucaria  de  Jussieu,  1789 
Section  bunya  Wilde  and  Eames,  1952 
Araucaria  brownii  sp.  nov. 

Plates  83-86 

Diagnosis.  Ovulate  cone,  4-5  x 5-0  cm  diameter,  pith  1-4  cm  diameter  near  point  of  attachment  to  peduncle, 
expanding  to  3-2  cm  wide  near  centre  of  cone.  Cortical  resin  canals  present.  Winged  cone-scales  1-7  cm  long  x 
1 • 1 cm  wide.  Bract  and  ovuliferous  scale  free  for  two-thirds  of  length,  both  with  a system  of  resin  canals. 
Ovules  0-8  cm  long  x 0-3  cm  wide,  wingless,  embedded  in  ovuliferous  scale  tissue  with  micropyle  oriented 
towards  cone  axis;  one  seed  per  cone-scale  complex.  Seed  integuments  with  prominent  branched  sclereids  of 
sclerotesta  arranged  in  a zig-zag  pattern.  Complex  system  of  vasculature  at  ovule  chalaza.  Nucellus  with  wavy 
apex  free  from  integuments  except  at  base,  0-2  mm  thick.  Megaspore  membrane  thin  (7  ^m)  and  discontinuous. 
Megagametophyte  composed  of  polygonal  cells  30-50  /xm  in  diameter. 

Holotype.  British  Museum  (Natural  History)  London,  V59205,  and  one  fragment  housed  at  Corfe  Castle 
Museum,  Dorset. 

Etymology.  This  cone  is  named  after  Mr.  P.  Anthony  Brown  of  Corfe  Castle,  Dorset  who  discovered  the 
specimen  and  made  it  available  for  study. 


EXPLANATION  OF  PLATE  83 

Figs.  1-8.  Araucaria  brownii  sp.  nov.  Holotype,  BMNH  V59205  from  Osmington  Mills,  Dorset,  b,  bract; 
end,  endotesta;  i,  integument;  Is,  ligular  sulcus;  m,  megagametophyte;  n,  nucellus;  os,  ovuliferous  scale; 
scl,  sclerotesta.  1 , cone  axis  region  showing  position  of  ovules,  and  arrangement  of  cone-scale  complexes, 
x 1.  2,  reverse  side  of  cone  in  fig.  1,  showing  cone  axis  region,  rhomboidal  cone-scale  complexes,  and 
flattened  nature  of  cone,  x 1.  3,  tangential  portion  showing  seed  transverse  sections,  xl.  4,  cone  portion 
with  part  of  axis  and  many  closely  spaced  cone-scale  complexes,  x 1 . 5,  reverse  side  of  portion  in  fig.  4,  show- 
ing limestone  nodule  matrix  and  numerous  cone-scale  complexes  with  ovules,  x 1.  6,  V59205  B 25.  Cone 
tangential  section  with  ovule  transverse  sections.  Note  large  calcite  crystals  replacing  most  bract  tissue,  x 7. 
7,  V59205  B 21,  longitudinal  section  of  cone-scale  showing  the  separation  of  bract  and  scale  resulting  in  a 
wide  ligular  sulcus,  x23.  8,  peel  of  V59205  B 6,  ovule  micropylar  end  showing  well-developed  seed 

integuments,  megagametophyte  tissue  and  wavy  nucellar  apex,  x 15. 


PLATE  83 


stockey,  Araucarian  cone 


660 


PALAEONTOLOGY,  VOLUME  23 


Description.  Cone  represented  by  four  pieces  which  are  slightly  flattened.  Plate  83,  figs.  1 and  2 show  both  sides 
of  central  portion  of  cone  revealing  position  of  cone  axis  and  numerous  helically  arranged,  seed-bearing  cone- 
scales.  Plate  83,  figs.  4 and  5 show  an  external  portion  of  the  cone.  This  piece  was  attached  to  that  shown  in 
PI.  83,  fig.  1 with  an  epoxy  before  sectioning.  Plate  83,  fig.  3 shows  a third  cone  fragment,  a tangential  cone 
piece  belonging  to  the  same  cone.  The  remaining  part  of  this  specimen  was  retained  by  Mr.  Brown  and  is 
housed  at  Corfe  Castle  Museum  in  Dorset. 

Cone  measures  4-5  x 5-0  cm  in  diameter  and  was  probably  spherical  in  shape  before  burial.  Pith  of  cone  axis 
reaches  a width  of  3-2  cm  and  approaches  1-4  cm  in  the  most  basal  portions.  Little  organic  material  pre- 
served in  central  part  of  cone,  and  no  peduncle  is  present  in  available  material.  Organic  remains  are  found  in 
some  cases  in  a region  corresponding  to  the  cortex  of  the  cone  axis  (PI.  86,  fig.  6).  Small  parenchymatous  cells  are 
often  found  in  groups  surrounding  resin  canals,  some  of  which  contain  a dark  opaque  substance  (PI.  86, 
fig.  6,  arrows).  A few  isolated  tracheids  have  been  identified  in  the  cone  axis  region;  these  exhibit  scalariform 
secondary  wall  thickenings.  Unfortunately  these  tracheids  are  isolated  and  their  exact  position  within  the  axis 
stele  is  questionable. 

Cone-scales  measure  1-7  cm  long  and  IT  cm  wide,  and  are  distinctly  winged.  On  first  examination  the  cone 
appears  similar  in  size,  shape,  and  appearance  to  those  of  the  genus  Agathis  Salisbury.  The  bract  and  scale, 
however,  are  free  for  most  of  their  length,  a character  typical  of  species  of  Araucaria  (PI.  83,  fig.  7).  Vascular 
system  of  the  ovuliferous  scales  consists  of  at  least  four  bundles  (PI.  84,  fig.  5).  Bract  apophyses  are  replaced  by 
coarsely  crystalline  calcite  making  vascular  bundles  generally  difficult  to  observe.  There  is  a system  of  resin 
canals  in  the  scale  as  well  as  the  bract,  although  their  number  and  placement  is  difficult  to  determine  because 
the  cone-scales  are  flattened. 

Ovules  of  A.  brownii  measure  0-8  cm  in  length  by  0-3  cm  in  diameter  (PI.  84,  figs.  1,  4).  One  wingless  seed 
per  cone-scale  complex  deeply  embedded  in  ovuliferous  scale  tissue  with  its  micropyle  oriented  towards  the  cone 
axis.  Although  both  Agathis  and  Araucaria  have  one  seed  per  scale,  only  seeds  of  Araucaria  are  wingless.  Ovules 
show  an  advanced  state  of  integumentary  development.  The  sarcotesta,  or  outer  layer,  is  represented  by  a thin 
layer  of  crushed  cells  20  pm  thick  (PI.  85,  fig.  1).  The  middle  sclerotesta  or  stony  layer  is  quite  thick  (up  to 
0-2  cm)  and  is  composed  of  thick-walled  branched  sclereids,  each  about  30  pm  in  diameter  (PI.  83,  fig.  8; 
PI.  84,  fig.  2;  PI.  85,  fig.  5).  These  cells  are  hexagonal  in  transverse  section  (PI.  85,  fig.  3)  with  very  small  lumens 
and  thick  walls.  In  many  cases  the  walls  are  no  longer  distinguishable,  but  the  entire  cell,  or  layer,  has  been 
replaced  by  calcite  (PI.  85,  figs.  1,  3).  The  endotesta,  or  inner  integumentary  layer,  is  thin,  up  to  three  cells 
in  thickness  and  often  crushed  (PI.  85,  fig.  1).  In  places  where  it  is  present,  the  cells  are  short  and  often  barrel- 
shaped with  relatively  thin  walls  (PI.  85,  fig.  2).  Integumentary  differentiation  within  the  ovules  indicates  a nearly 
mature  developmental  stage. 

The  non-adnate  nature  of  the  nucellus  and  integument  is  shown  in  all  of  the  ovules  examined  (PI.  83,  fig.  8; 
PI.  84,  fig.  7).  Where  well  preserved  the  nucellus  has  a thick  cuticle  (PI.  83,  figs.  6,  8;  PI.  84,  figs.  3,  6; 
PI.  86,  figs.  1,  2).  It  is  basally  attached  to  the  inner  integumentary  layer  and  appears  somewhat  shrunken 
(PI.  86,  fig.  2).  The  nucellar  apex  appears  convoluted  as  in  living  araucarians  (PI.  83,  fig.  8).  In  extant  plants 
it  protrudes  out  of  the  micropyle  at  the  time  of  pollination  and  later  may  retract  by  drying  or  by  subsequent 
integumentary  growth  (Eames  1913).  Some  authors  (Darrow  1936;  Eames  1913)  have  suggested  that  the  con- 
voluted apex  was  the  result  of  pollen  tube  damage;  however,  on  examining  some  living  araucarian  ovules,  a 
disruption  of  the  apex  by  pollen  tube  damage  seems  unlikely.  Plate  85,  fig.  4 shows  the  wavy  nucellar  apex  of  the 
extant  A.  montana  Brongn.  et  Gris,  and  is  typical  of  most  known  araucarians  at  this  stage  of  development.  The 
configuration  of  the  apex  appears  to  represent  a drying  phenomenon  rather  than  the  result  of  extensive  pollen 


EXPLANATION  OF  PLATE  84 

Figs.  1-7.  Araucaria  brownii  sp.  nov.  Holotype,  BMNH  V59205.  b,  bract;  i,  integument;  m,  megagameto- 
phyte;  n,  nucellus;  os,  ovuliferous  scale;  s,  seed.  1 , isolated  seed,  x 1 0.  2,  paradermal  section  of  sclerotesta 

cells  showing  zig-zag  cell  arrangement.  Arrows  indicate  branched  sclereids,  x 380.  3,  isolated  nucellus  with 

apex  removed,  x 17.  4,  V59205  B 6,  cone  longitudinal  section  showing  sunken  nature  of  seeds  within  the 
cone-scale  complex,  x 7.  5,  V59205  B 25  transverse  section  of  ovule  showing  lateral  vascular  bundle  of 

ovuliferous  scale  outside  of  the  seed  integuments,  x 37.  6,  V59205  A 40,  ovule  transverse  section  showing 

thick  wavy  nucellus  and  cellular  megagametophyte  tissue,  x 90.  7,  V59205  B 26,  ovule  transverse  section 

showing  the  relationship  of  integument,  nucellus,  and  cellular  megagametophyte,  x 85. 


PLATE  84 


stockey,  Araucarian  cone 


662 


PALAEONTOLOGY,  VOLUME  23 


tube  damage.  The  ease  with  which  the  nucellus  is  removed  from  isolated  ovules  is  due  to  its  narrow  attachment 
as  well  as  its  thick  cuticle  (PI.  84,  fig.  3).  External  surface  shows  little  cellular  detail  while  internally  outlines  of 
elongate,  rectangular  cells  are  visible  (PI.  85,  fig.  7). 

Ovule  attachment  and  vascularization  are  difficult  to  determine  with  peels.  However,  thin  sections  of  the 
chalaza  of  some  ovules  where  preservation  of  the  bract  and  ovuliferous  scale  is  partial  show  a well-developed 
system  of  conducting  cells  (PI.  86,  fig.  5).  Using  scanning  electron  microscopy,  the  chalazal  end  of  each  ovule 
exhibits  a series  of  small  holes  penetrating  the  sclerotesta  (PI.  84,  fig.  6,  arrows),  as  in  the  living  A.  bidwillii 
Hooker.  There  appears  to  be  what  Wilde  and  Eames  (1948,  p.  326)  term  a vascular  ‘plexus’,  a complex  system  of 
vascularization  near  the  ovule  base.  Holes  in  the  mature  seed  integuments  correspond  to  points  of  entry  of 
numerous  vascular  bundles.  The  same  type  of  attachment  also  occurs  in  A.  mirabilis  from  the  Jurassic  Cerro 
Cuadrado  Petrified  Forest  (Stockey  1975). 

Most  ovules  show  some  tissue  remains  inside  the  nucellar  cavity.  The  megaspore  membrane  of  some  ovules 
is  thin  (7  fin l)  (PI.  84,  fig.  7),  and  similar  to  that  in  living  araucarian  cones  at  a comparable  stage  of 
development  (Eames  1913;  Burlingame  1915;  Thomson  19056).  Many  of  the  ovules  reveal  preservation  of  tissues 
within  this  membrane.  In  most,  megagametophyte  tissue  is  either  poorly  preserved  or  represented  by  a free 
nuclear  stage  of  development  at  the  time  of  preservation  (PI.  84,  fig.  8;  PI.  86,  figs.  2,  7).  Other  ovules  show 
cellular  preservation  of  the  megagametophyte  (PI.  84,  fig.  6;  PI.  86,  fig.  4).  These  polygonal  cells  (30-50  fim  in 
diameter)  occur  in  the  outer  portions  of  the  megagametophyte  proper.  No  ovules  have  been  found  with  solid 
megagametophyte  tissue  preserved  within  the  seed  cavity.  However,  some  ovules  do  show  two  distinct  regions 
of  poorly  preserved  tissue  (PI.  86,  fig.  2).  The  boundary  between  the  two  regions  appears  to  be  a discontinuous 
layer.  In  other  ovules  (PI.  86,  fig.  4)  the  cellular  megagametophyte  and  a centrally  located  region  probably  repre- 
senting the  embryo  are  replaced  by  calcite.  Other  specimens  contain  a four-parted  cellular  structure  that  may 
represent  an  embryo  with  four  cotyledons  (PI.  86,  fig.  3,  arrows).  An  alternate  possibility  is  that  this  may  have 
been  a partially  formed  megagametophyte  at  the  time  of  preservation.  The  seed  itself  is  not  crushed,  even 
though  parts  of  the  integuments  are  very  crumbly. 


DISCUSSION 

The  spherical  shape  of  the  Osmington  Mills  cone  with  helically  arranged  cone-scales,  large  pith  in  the 
cone  axis  region,  cortical  resin  canals,  and  presence  of  one  seed  per  ovuliferous  scale  are  general 
characteristics  of  cones  from  the  family  Araucariaceae.  The  presence  of  a ligular  sulcus  (space 
between  the  ovuliferous  scale  tip  and  the  bract)  and  wingless  seeds  suggest  affinities  with  the  genus 
Araucaria.  The  cone  of  A.  brownii  was  apparently  in  a relatively  mature  state  of  development  at  the 
time  of  fossilization.  The  appearance  of  the  three-layered  integument  with  a thin  sarcotesta,  thick 
sclerotesta  composed  of  elongate  branched  sclereids,  and  a thin  layer  of  endotesta  composed  of  thin- 
walled  cells  is  characteristic  of  araucarian  cones  near  maturity  (Eames  1913;  Burlingame  1915;  Wilde 
and  Eames  1948;  Stockey  1978).  The  seed  contents  including  the  configuration  of  the  nucellus  and 
thin  megaspore  membrane  support  this  view.  A cellular  megagametophyte  is  present  with  a hollow 
central  cavity  (PI.  86,  figs.  2,  4)  that  probably  represents  the  remains  of  an  embryo  rather  than  free 
nuclear  megagametophyte.  This  embryo  may  have  aborted  or  more  likely  deteriorated  prior  to 
preservation  since  it  is  likely  that  the  cone  remained  floating  in  water  for  some  time  prior  to  its 


EXPLANATION  OF  PLATE  85 

Figs.  1 -7.  Araucaria  brownii  sp.  nov.  Holotype,  BMNH  V59205,  scanning  electron  micrographs,  end,  endo- 
testa; n,  nucellus;  sar,  sarcotesta;  scl,  sclerotesta.  1 , ovule  transverse  section  showing  three  integumentary 
layers  at  a late  developmental  stage,  x 90.  2,  cells  of  the  endotesta,  x 425.  3,  transverse  section  of  integument 
showing  hexagonal  sclerotesta  cells  completely  replaced  by  calcite,  x 400.  4,  A.  montana  Brongn.  et  Gris., 
nucellar  apex  from  a mature  seed,  x 100.  5,  surface  view  of  sclerotesta  with  sarcotesta  removed  showing 

elongate  interlocking  sclereids,  x 425.  6,  seed  chalaza,  surface  of  sclerotesta.  Arrows  indicate  holes  of  the 

ovular  vascular  bundles  in  the  plexus,  x 90.  7,  elongate  nucellar  cells,  x 450. 


PLATE  85 


stockey,  Araucarian  cone 


664 


PALAEONTOLOGY,  VOLUME  23 


burial.  Plate  86,  fig.  3 may  show  the  cellular  remains  of  such  an  embryo.  Extant  araucarian  cones, 
at  the  time  when  free  nuclear  megagametophyte  is  present  (about  the  time  of  pollination),  show 
ovules  with  the  integumentary  layers  of  approximately  equal  thickness  with  little  if  any  expansion 
and  thickening  of  the  sclerotesta  (Wilde  and  Eames  1948;  Stockey  1978). 

According  to  Wilde  and  Eames  (1952)  the  living  genus  Araucaria  may  be  divided  into  four  sections: 
Columbea , Bunya,  Eutacta,  and  Intermedia  based  on  seedling  morphology,  foliage  type,  and  cone 
morphology.  The  differences  in  cone  structure  between  the  section  Columbea  and  the  other  three 
sections  is  distinct,  while  the  differences  between  the  other  three  are  more  subtle.  The  Columbea 
species  found  only  in  South  America  have  wingless  cone-scales,  the  bract  and  scale  are  nearly  com- 
pletely fused  and  never  separate  after  the  scales  are  shed  from  the  cone  axis.  Sections  Eutacta, 
Intermedia,  and  Bunya  have  winged  cone-scales,  those  of  Intermedia  being  widest,  up  to  10  cm  in 
A.  klinkii  Lauterb.  (White  1947).  Seeds  in  the  closely  related  Eutacta  and  Intermedia  sections  are 
never  removed  from  the  tightly  fused  bract  and  scale.  The  winged  cone-scale  complex  is  the  unit  of 
dispersal  for  these  species.  Seeds  from  A.  bidwillii,  the  only  living  member  of  the  section  Bunya,  are 
easily  removed  from  the  cone-scales  which  are  more  fleshy  than  those  of  the  Eutacta  and  Intermedia 
sections.  These  seeds  are  often  dispersed  by  birds  and  small  animals  in  Queensland  but  are  easily 
removed  from  the  cone-scale  complex  even  though  the  cone  disaggregates  as  in  the  other  species. 

A.  mirabilis  from  the  Jurassic  Cerro  Cuadrado  Petrified  Forest  is  also  included  within  the  Section 
Bunya  (Calder  1953;  Stockey  1975, 1978).  These  ovulate  cones  are  only  one-third  as  large  as  those  of 
A.  bidwillii  at  maturity  and  have  winged  cone-scales,  a zig-zag  pattern  of  sclereids  in  the  sclerotesta 
of  the  seed  integument,  a complex  system  of  vasculature  at  the  seed  chalaza,  and  a dicotyledonous 
embryo  characteristic  of  A.  bidwillii  (Wilde  and  Eames  1948;  Stockey  1975,  1978).  There  is  some 
evidence  to  suggest  that  these  cones  shed  their  seeds  and  not  their  scales  at  maturity  (Stockey 
1978). 

Another  ovulate  cone  which  can  be  considered  to  be  closely  related  to  these  three  Bunya  species  is 
Araucarites  bindrabunensis  (Vishnu-Mittre  1954)  from  the  Jurassic  of  India  which  shows  a slightly 
larger  size  than  most  Araucaria  mirabilis  cones,  and  is  also  larger  than  the  cone  from  Osmington 
Mills.  The  origin  of  the  cone-scale  complex  vascular  supply,  nature  of  the  winged  cone-scales, 
presence  of  a ligular  sulcus,  and  vascularization  of  the  ovuliferous  scale  tip  (ligule)  place  it  in  the 
section  Bunya. 

The  cone  described  here  from  Osmington  Mills  in  Dorset  should  also  be  considered  under  the 
section  Bunya  of  the  genus  Araucaria.  The  winged  cone-scales,  zig-zag  sclereid  pattern  and  vascular 
plexus,  and  deep  ligular  sulcus  are  similar  to  A.  mirabilis  and  A.  bidwillii.  These  comparisons  extend 
the  range  of  the  section  Bunya  into  the  Northern  Hemisphere  where  it  was  probably  widespread 
during  the  Jurassic  and  Cretaceous. 

Acknowledgements.  I thank  Dr.  C.  R.  Hill,  British  Museum  (Natural  History),  and  Mr.  P.  A.  Brown  for  making 
the  material  available  for  study;  and  Dr.  T.  N.  Taylor  for  review  of  the  manuscript  and  use  of  laboratory 
facilities.  This  publication  is  dedicated  to  the  memory  of  the  late  James  M.  Schopf.  I acknowledge  NSERCC 
Grant  A-6908. 


EXPLANATION  OF  PLATE  86 

Figs.  1-7.  Araucaria  brownii  sp.  nov.  Holotype,  BMNH  V59205.  e,  embryo;  i,  integument;  m,  megagameto- 
phyte; n,  nucellus;  p,  plexus.  1,  V59205  B 24,  ovule  transverse  section  showing  lateral  extensions  of  the  seed 
integuments  and  well-preserved  endotesta  cells,  x9.  2,  V59205  A 40,  seed  transverse  section  with  well- 
preserved  nucellus,  megagametophyte,  and  a possible  embryo,  x 25.  3,  V59205  B 23,  transverse  section 

of  seed  showing  possible  embryo  with  four  cotyledons,  x 35.  4,  Y59205  B 25,  transverse  section  of  seed  with 
cellular  megagametophyte  and  crystalline  central  area,  possibly  representing  an  embryo,  x 50.  5,  V59205 

B 26,  transverse  section  near  seed  chalaza  showing  vascular  plexus  leading  into  ovule,  x 15.  6,  V59205 

C 8,  longitudinal  section  of  resin  canal  (arrows)  in  cortex  of  cone  axis,  x 30.  7,  transverse  section  of  ovule 
showing  disorganized  megagametophyte  tissue,  x 925. 


PLATE  86 


stockey,  Araucarian  cone 


666 


PALAEONTOLOGY,  VOLUME  23 


REFERENCES 

arkell,  w.  J.  1947.  Geology  of  the  Country  around  Weymouth,  Swanage,  Corfe  and Lulworth.  Mem.  Geol.  Survey 
of  Great  Britain,  1-386. 

Burlingame,  L.  L.  1 9 1 3.  The  morphology  of  Araucaria  brasiliensis.  I.  The  staminate  cone  and  male  gametophyte. 
Bot.  Gaz.  55,  97-114. 

- — 1914.  The  morphology  of  Araucaria  brasiliensis.  II.  The  ovulate  cone  and  female  gametophyte.  Ibid.  57, 
490-508. 

— 1915.  The  morphology  of  Araucaria  brasiliensis.  Fertilization,  the  embryo  and  the  seed.  Ibid.  59,  1-38. 
calder,  M.  G.  1953.  A coniferous  petrified  forest  in  Patagonia.  Bull.  Brit.  Mus.  (Nat.  Hist.)  Geol.  2,  99-138. 
chamberlain,  c.  j.  1935.  Gymnosperms:  Structure  and  Evolution.  Univ.  Chicago  Press. 
darrow,  B.  s.  1936.  A fossil  araucarian  embryo  from  the  Cerro  Cuadrado  of  Patagonia.  Bot.  Gaz.  98,  328-337. 
eames,  A.  J.  1913.  The  morphology  of  Agathis  australis.  Ann.  Bot.  27,  191-204. 

hirmer,  M.  1936.  Die  Bliiten  der  Coniferen.  I.  Entwicklungsgeschichte  und  Vergleichende  Morphologie  des 
weiblichen  Bliitenzapfens  der  Coniferen.  Biblio.  Bot.  23,  1-100. 
joy,  K.  w.,  willis,  A.  j.  and  lacey,  w.  s.  1956.  A rapid  cellulose  peel  technique  in  palaeobotany.  Ann.  Bot. 
(n.s.),  20,  635-637. 

kendall,  m.  w.  1949.  A Jurassic  member  of  the  Araucariaceae.  Ibid.  13,  151-161. 

robison,  c.  r.  and  miller,  c.  n.  jr.  1975.  Glycol  methacrylate  as  an  embedding  medium  for  lignitic  plant 
fossils.  J.  Paleont.  49,  559-561. 

seward,  A.  c.  and  ford,  s.  o.  1906.  The  Araucarieae,  recent  and  extinct.  Phil.  Trans.,  London  ( B ),  198, 
305-411. 

stockey,  R.  A.  1975.  Seeds  and  embryos  of  Araucaria  mirabilis.  Am.  J.  Bot.  62,  856-868. 

— 1977.  Reproductive  biology  of  the  Cerro  Cuadrado  (Jurassic)  fossil  conifers:  Par  araucaria  patagonica.  Ibid. 
64,  733-744. 

— 1978.  Reproductive  biology  of  Cerro  Cuadrado  fossil  conifers:  ontogeny  and  reproductive  strategies  in 
Araucaria  mirabilis  (Spegazzini)  Windhausen.  Palaeontographica,  B 166,  1-15. 

Thomson,  R.  B.  1905a.  Preliminary  note  on  the  Araucarineae.  Science  (N.S.),  22,  88. 

— 19056.  The  megaspore-membrane  of  the  gymnosperms.  Univ.  Toronto  Studies  Bio.  Ser.  4,  85-146. 

— 1907.  The  Araucarieae— A ‘Proto-Siphonogamic’  method  of  fertilization.  Science  (n.s.),  25,  271-272. 
— 1913.  On  the  comparative  anatomy  and  affinities  of  the  Araucarineae.  Phil.  Trans.  R.  Soc.  Lond.  (B), 
204,  1-50. 

white,  c.  t.  1947.  Notes  on  two  species  of  Araucaria  in  New  Guinea,  and  a proposed  new  section  of  the 
genus.  J.  Arnold  Arboretum,  28,  259-260. 

wieland,  G.  R.  1935.  The  Cerro  Cuadrado  Petrified  Forest.  Publ.  Carnegie  Inst.,  Washington.  449  pp. 
wilde,  m.  h.  and  eames,  a.  j.  1948.  The  ovule  and  ‘seed’  of  Araucaria  bidwilli  with  discussion  of  the  taxonomy 
of  the  genus.  I.  Morphology.  Ann.  Bot.  (N.S.),  12,  311-326. 

— 1952.  The  ovule  and  ‘seed’  of  Araucaria  bidwilli  with  discussion  of  the  taxonomy  of  the  genus. 
II.  Taxonomy.  Ibid.  16,  27-47. 

DR.  R.  A.  STOCKEY 
Department  of  Botany 
The  University  of  Alberta 
Edmonton,  Alberta 


Typescript  received  24  June  1979 

Revised  typescript  received  20  November  1979 


NOMENCLATURE  AND  HOMOLOGY  IN 
PERIDINIALEAN  DINOFLAGELLATE 
PLATE  PATTERNS 

by  GEOFFREY  L.  EATON 


Abstract.  The  apical  and  antapical  series  of  peridinialean  dinoflagellate  thecal  plates  are  redefined  relative  to 
the  cingulum.  They  are  then  compatible  with  the  Kofoidian  pre-  and  postcingular  series,  and  the  need  to 
recognize  anterior  and  posterior  intercalary  series  is  removed.  The  concept  of  apical  closing  and  antapical 
closing  series  is  introduced.  Homologous  and  corresponding  plates  are  recognized  in  fifteen  selected  modern  and 
fossil  dinoflagellates  by  comparing  interseries  relationships  with  respect  to  a model  plate  pattern.  The  differences 
between  the  selected  patterns  are  due  to  three  variable  effects.  First,  the  reduction  in  plate  number  through 
simplification,  where  one  plate  in  one  pattern  corresponds  to  two  or  more  plates  in  another  pattern.  This 
critically  affects  interseries  relationships.  Secondly,  the  primary  development  of  fewer  plates  without  affecting 
interseries  relationships.  Thirdly,  the  variation  in  the  relative  size  of  certain  plates.  The  interaction  of  these  three 
effects  resulted  in  the  comparatively  independent  evolution  of  epithecae  and  hypothecae.  Reduction  in  over-all 
plate  number,  particularly  through  the  primary  development  of  fewer  plates,  may  well  represent  a fundamental 
trend  in  the  evolution  of  peridinialean  plate  patterns. 


The  dinoflagellates  of  the  Order  Peridiniales  Haeckel  1894  are  often  informally  described  as 
‘armoured’.  They  are  so  called  because  their  cell  covering  includes  a layer  of  rigid,  polygonal, 
suturally  united,  cellulosic  plates,  termed  the  theca.  Text-fig.  1 shows  thecal  morphology  and 
nomenclature  in  two  typical  peridinialean  dinoflagellates,  Protoperidinium  depressum  (Bailey)  Balech 
1 974  and  Gonyaulax  spinifera  (Claparede  and  Lachmann)  Diesing  1866.  The  theca  is  divided  into  two 
parts,  epitheca  (anterior)  and  hypotheca  (posterior),  which  are  separated  by  an  equatorial  groove 
termed  the  cingulum.  The  two  ends  of  the  cingulum  are  separated  on  the  ventral  surface  by  a more  or 
less  longitudinal  groove  termed  the  sulcus.  Motility  is  achieved  by  the  beating  of  two  flagella  (not 
shown  in  text-fig.  1)  which  originate  from  the  sulcus.  The  transverse  flagellum  lies  within  the 
cingulum,  while  the  longitudinal  flagellum  lies  within,  and  extends  posteriorly  beyond,  the  sulcus.  The 
thecal  plates  are  arranged  in  roughly  parallel  transverse  series.  Differences  in  tabulation;  that  is  the 
number,  shape,  and  arrangement  (plate  pattern)  of  the  thecal  plates,  have  long  been  used  as  the  main 
criterion  for  taxonomic  separation  within  the  Peridiniales.  The  accepted  system  of  thecal  plate 
nomenclature  was  defined  by  Kofoid  (1907,  1909,  1911). 

The  fossil  peridinialean  dinoflagellate  record  ranges  back  at  least  200  million  years  into  the  Late 
Triassic  period.  However,  in  terms  of  representing  the  absolute  geological  history  of  the  Peridiniales, 
this  record  has  only  limited  effectiveness.  This  is  because  all  fossilized  dinoflagellates  attributed  to  the 
Peridiniales  are  non-motile  cysts  rather  than  motile  thecae,  and  modern  studies  show  that  only  a very 
small  proportion  of  living  peridinialeans  produce  potentially  fossilizable  cysts.  Comparisons 
between  modern  thecae  and  fossil  cysts  show  that  not  all  modern  plate  patterns  have  been  recognized 
in  the  fossil  record,  and  some  fossil  plate  patterns  are  unknown  in  modern  dinoflagellates.  Accepting 
the  limitations  of  the  fossil  record,  and  the  fact  that  cysts  only  rarely  show  full  details  of  their  parent 
thecal  tabulation,  it  is  still  possible  that  the  relative  distribution  of  the  different  plate  patterns  through 
geological  time  may  provide  some  evidence  of  trends  in  plate  pattern  evolution.  The  recognition  of 
such  trends  is  dependent  on  the  critical  assessment  of  the  similarities  and  differences  between  different 
plate  patterns.  Such  an  assessment  will  involve  the  recognition  of  homologous  plates  in  different 
patterns.  In  my  own  studies  on  fossil  dinoflagellates  I have  found  that  a strict  application  of  Kofoid’s 


[Palaeontology,  Vol.  23,  Part  3, 1980,  pp.  667-688.| 


668 


PALAEONTOLOGY,  VOLUME  23 


plate  nomenclature  often  results  in  apparently  homologous  plates  in  different  patterns,  being 
assigned  to  different  transverse  plate  series.  I believe  that  compatibility  between  nomenclature  and 
homology  is  essential  for  the  recognition  of  evolutionary  trends,  and  that  it  can  only  be  achieved  by 
modifying  certain  aspects  of  Kofoid’s  system.  Discussion  of  the  need  for  this  modification  and  a way 
of  effecting  it,  forms  the  basis  of  this  paper. 

According  to  Evitt  et  al.  (1976)  fossil  cyst  plate  patterns  should  be  discussed  in  terms  of  their 
paratabulatory  nomenclature  (paraplates,  parasutures,  etc.).  However,  in  this  paper  on  modern 


VENTRAL  SURFACE  DORSAL  SURFACE 

text-fig.  1 . Thecal  morphology  of  two  modern  peridinialean  dinoflagellates.  Upper,  Protoperidinium  depression 
(Bailey)  Balech  1974.  Lower,  Gonyaulax  spinifera  (Claparede  and  Lachmann)  Diesing  1866.  Interpretation  of 
the  transverse  plate  series  is  conventional  Kofoidian.  The  distribution  and  number  of  cingular  and  sulcal  plates 
in  P.  depressum  is  assumed  to  be  typical  of  the  genus.  In  G.  spinifera  only  the  anterior  (a.s.)  and  posterior  (p.s.) 

sulcal  plates  are  annotated. 


EATON:  DINOFLAGELLATE  PLATE  PATTERNS 


669 


thecae  and  fossil  cysts  I wish  to  avoid  the  use  of  a dual  ‘tabulation/paratabulation’  nomenclature. 
Therefore  I assume  that  the  fossil  cyst  paraplate  patterns  are  a fair  representation  of  their  parent 
thecal  plate  patterns,  and  treat  them  all,  modern  and  fossil,  simply  as  plate  patterns. 


ORIGIN  AND  DEVELOPMENT  OF  THE  KOFOID  SYSTEM  OF 
THECAL  PLATE  NOMENCLATURE 

Although  the  system  of  peridinialean  thecal  plate  nomenclature  which  has  been  generally  used  for  the 
past  seventy  years  is  attributed  to  Kofoid,  it  should  be  remembered  that  he  was  clearly  influenced  by 
several  nomenclatural  systems  proposed  by  earlier  workers,  e.g.  Stein,  Biitschli,  Schiitt,  Paulsen, 
Faure-Fremiet  (see  Kofoid  1909,  p.  44).  All  these  earlier  workers  recognized  that  thecal  plates  are 
arranged  in  transverse  rows,  and  that  there  are  four  major  plate  series,  two  anterior  to  the  equator 
and  two  posterior  to  the  equator.  Various  names  had  been  applied  to  these  series  (see  Kofoid  1909,  p. 
44),  but  those  used  by  Biitschli  (1885)  were  closest  to  Kofoid’s  subsequent  terminology.  Biitschli 
described  the  most  anteriorly  positioned  series  as  apical,  and  the  most  posteriorly  positioned  as 
antapical.  The  two  intervening  series  were  termed  pre-equatorial  (anterior)  and  post-equatorial 
(posterior). 

Kofoid  recognized  seven  transverse  plate  series,  comprising  the  four  major  series  plus  the  cingular 
series  and  two  incomplete  intercalary  series.  Each  series  was  designated  by  superscript  acute  accent 
marks,  figures  or  letters,  or  simply  by  letters.  The  series  were  named  from  apex  to  antapex  as:  apical 
('),  anterior  intercalary  (a),  precingular  ("),  cingular  (c),  postcingular  ("'),  posterior  intercalary  (p), 
antapical  ("").  The  plates  in  each  series  were  numbered  in  sequence,  anticlockwise  (in  apical  view) 
from  the  ventral  surface.  Additional  plates  at  the  extreme  apex  or  within  the  sulcus  were  individually 
designated,  e.g.  apical  closing  plate  (cl.  pi.).  During  subsequent  use,  Kofoid’s  system  has  remained 
unchanged  except  for  the  designation  of  the  sulcal  plates  (s)  and  the  use  of  various  abbreviations  to 
designate  additional  individual  plates.  The  typical  application  of  Kofoid’s  nomenclature  to  P.  depres- 
sion and  G.  spinifera  is  shown  in  text-fig.  1 . These  two  forms  together  illustrate  all  seven  of  Kofoid’s 
transverse  plate  series.  Their  respective  tabulation  formulae  are:  4',  3a,  7",  3c,  5"',  Op,  2"",  6-7s,  and 
1 ap.  cl.,  4',  Oa,  6”,  6c,  6"',  lp,  1 5s. 

The  most  extensive  discussion  of  thecal  plate  nomenclature  is  given  in  Kofoid  (1909,  pp.  40-45), 
but  definitions  of  the  various  plate  series  are  also  found  in  Kofoid  (1907,  1911). 

Kofoid  (1907,  p.  179)  defined  the  four  major  plate  series  with  reference  to  modern  Ceratium 
Schrank  1793.  Kofoid  stated:  ‘I  shall  use  the  term  apical  for  the  anterior  series  of  plates  only,  and 
shall  designate  the  series  anterior  to  and  contiguous  to  the  girdle  [cingulum]  as  precingular  (prec.), 
and  that  posterior  to  and  contiguous  to  it  as  postcingular  (postc.)  and  the  posterior  ones  as  antapicals 
(antap.).’ 

Kofoid  (1909,  pp.  26-28)  next  applied  his  nomenclature  to  modern  peridiniacean  dinoflagellates, 
using  P.  steini  (Jorgensen)  as  an  example.  His  nomenclatural  interpretation  of  P.  steini  is  equally 
applicable  to  P.  depressum  (text-fig.  1).  Kofoid  interpreted  the  apical  plates  as  ‘those  whose  apical 
ends  border  the  apical  pore’  (ap.  po.  in  text-fig.  1).  The  combination  of  this  interpretation  of  the 
apicals  and  Kofoid’s  earlier  interpretation  of  the  precingulars  leaves  three  plates  unaccounted  for  on 
the  dorsal  surface.  These  plates  ‘intercalate’  between  the  apicals  and  precingulars  and  were  referred  to 
the  anterior  intercalary  series  (la-3a),  a term  Kofoid  had  previously  used  in  his  original  description 
of  Heterodinium  Kofoid  1906. 

In  his  studies  on  modern  Gonyaulax  Diesing  1866,  Kofoid  (1911,  p.  194)  interpreted  the  apical 
plates  as  ‘those  in  contact  with  the  apex’.  He  recognized  that  in  this  genus  the  apex  does  not  have  an 
open  pore,  but  is  occupied  by  a small  apical  closing  plate  ( 1 ap.  cl.  in  text-fig.  1 ).  Kofoid  designated  as 
anterior  intercalary  those  plates  anterior  to  the  precingular  series  but  not  in  contact  with  the  apex. 
This  series  was  not  recognized  in  all  species  of  Gonyaulax.  Kofoid  also  introduced  the  concept  of 
a posterior  intercalary  series  with  reference  to  Gonyaulax.  The  single  plate  (lp)  assigned  to  this 
series  lies  posterior  to  postcingulars  Y"  and  2"',  and  anterior  to  antapical  V"  which  occupies  the 
antapex. 


670 


PALAEONTOLOGY,  VOLUME  23 


Kofoid  realized  that  thecal  plates  are  arranged  in  rows  roughly  parallel  to  the  cingulum.  This  led 
him  to  use  the  cingulum  rather  than  the  geometric  equator  as  a basis  for  defining  transverse  plate 
series.  This  approach  recognized  the  fundamental  importance  of  the  structure  which  divides  the  theca 
into  epitheca  and  hypotheca.  Kofoid  (1909,  p.  43)  believed  that  his  recognition  of  transverse  series 
throughout  the  theca  clarified  the  confused  situation  that  had  previously  existed  over  the 
nomenclature  of  plates  anterior  to  the  precingulars.  His  beliefs  would  seem  to  have  been  justified 
by  the  subsequent  application  of  his  nomenclature  to  modern  dinoflagellates  and  to  fossil  forms 
ranging  back  to  the  Triassic  period. 


C.  HIRUND1NELLA  P.  DEPRESSUM  G.  SPINIFERA 

text-fig.  2.  Conventional  Kofoidian  interpretation  of  tabulation  in  polar  views  of  Ceratium  hirundinella 
(Muller)  Schrank  1793,  Protoperidinium  depressum  (Bailey)  Balech  1974,  Gonyaulax  spinifera  (Claparede  and 
Lachmann)  Diesing  1866.  Upper,  epithecae.  Lower,  hypothecae. 


A problem  in  designating  apparently  homologous  plates 

Although  Kofoid  did  not  recognize  any  intercalary  plates  in  Ceratium,  a case  can  sometimes  be  made 
for  a single  anterior  intercalary  in  Ceratium  hirundinella  (Muller)  Schrank  1793  (text-fig.  2).  In  a 
particular  form  of  this  species.  Wall  and  Evitt  (1975,  p.  21)  designated  as  apical  4'  a plate  which  does 
not  reach  the  tip  of  the  apical  horn.  They  admitted  that  strictly  speaking  this  plate  should  be 
designated  anterior  intercalary,  but  to  do  so  would  only  lead  to  confusion.  They  argued  that  since  the 
homology  of  this  plate  and  the  fourth  apical  of  other  species  of  Ceratium  is  so  obvious,  it  is  better  to 
consider  this  plate  as  a shortened  apical.  In  this  particular  case  Wall  and  Evitt  considered  the 
recognition  of ‘obvious’  homology  to  be  more  important  than  the  strict  application  of  a definition  or 
rule. 

This  approach  can  also  be  applied  to  other  dinoflagellates,  for  instance,  the  fossil  taxon 
Hystrichogonyaulax  cladophora  (Deflandre)  Stover  and  Evitt  1978  and  certain  species  of  fossil 


EATON:  DINOFLAGELLATE  PLATE  PATTERNS 


671 


Ctenidodinium  Deflandre  1938.  In  some  well-preserved  specimens  of  H.  cladophora  (text-fig.  3)  a plate 
can  be  recognized  anterior  and  adjacent  to  3"  and  4".  This  plate  does  not  touch  the  apical  closing 
plate  (1  ap.  cl.)  and  is  therefore  designated  anterior  intercalary  la.  Two  such  plates  are  recognizable 
in  Ctenidodinium  pachydermum  (Deflandre)  Gocht  (1970,  pi.  29,  fig.  5)  and  Ctenidodinium  sp.  (text- 
fig.  3),  and  2a  in  this  pattern  appears  to  be  homologous  with  la  in  H.  cladophora.  Also,  2a  in 
Ctenidodinium  sp.  and  la  in  H.  cladophora  appear  to  be  homologous  with  apical  3'  in  Gonyaulax 
polyedra  Stein  1883  (text-fig.  3).  If  this  interpretation  of  homology  is  correct,  then  this  particular 
anterior  intercalary  plate  in  the  two  fossil  taxa  could  be  interpreted  as  a shortened  apical. 


text-fig.  3.  Conventional  Kofoidian  interpretation  of  epithecal  tabulation  in  Hystrichogonyaulax  cladophora 
(Deflandre)  Stover  and  Evitt  1978,  Ctenidodinium  Deflandre  1938  sp.,  Gonyaulax  polyedra  Stein  1883. 


Kofoid’s  (1911,  pp.  194-195)  own  comments  on  the  anterior  intercalaries  in  gonyaulacacean 
dinoflagellates  are  significant  here.  Kofoid  designated  as  anterior  intercalary,  plates  in  the  apical 
region  which  are  ‘crowded  away  from  contact  with  the  apex  ...  as  well  as  other  plates  lying  between 
the  apical  and  precingular  series’.  He  also  considered  the  two  anterior  intercalaries  lying  laterally  and 
ventrally  to  the  right  of  the  greatly  reduced  apical  4'  in  G.  polyedra  (text-fig.  3)  to  be  plates  which  had 
been  ‘crowded  away’  from  the  apex.  Kofoid  remarked  further  (Kofoid  191 1,  p.  239)  that  the  area  of 
la  had  probably  ‘split  off’  from  the  edge  of  apical  4',  and  he  also  illustrated  one  specimen  of  G. 
polyedra  (Kofoid  191 1 , pi.  14,  fig.  29)  in  which  intercalary  2a  actually  touches  the  apical  closing  plate. 
There  is  no  doubt  that  Kofoid  considered  intercalaries  la  and  2a  in  G.  polyedra  to  be  territorially 
apical,  but  their  spatial  relationship  with  the  extreme  apex  required  their  designation  as  anterior 
intercalary. 

The  anterior  intercalary  plates  in  H.  cladophora  and  Ctenidodinium  sp.  seem  to  be  apicals  which 
have  been  shortened  and  crowded  away  from  the  apex,  and  according  to  Kofoid’s  comments  on  G. 
polyedra  their  designation  as  intercalary  is  entirely  justified.  Also,  it  can  be  argued  that  Wall  and  Evitt 
should  have  adopted  this  approach  with  C.  hirundinella.  Apical  4'  could  be  interpreted  as  being 
crowded  away  from  the  apex  to  occupy  an  anterior  intercalary  position,  and  this  plate  could  then  be 
designated  la.  This  would  not  lead  to  the  confusion  Wall  and  Evitt  suggested.  It  would  simply  reflect 
the  strict  application  of  a universally  recognized  rule,  and  any  discussion  of  homologous 
relationships  with  the  apical  plates  of  other  taxa  would  be  of  secondary  importance.  However, 
against  this  it  can  be  argued  that  the  recognition  of  homologous  plates  in  different  dinoflagellates  is  in 
fact  of  primary  importance,  and  is  critical  to  the  understanding  of  the  evolution  of  thecal  plate 
patterns.  Therefore,  since  the  Kofoid  rules  require  that  apparently  homologous  plates  in  different 
taxa  are  assigned  to  different  plate  series,  Kofoid’s  method  of  defining  these  series  should  be 
re-evaluated. 


672 


PALAEONTOLOGY,  VOLUME  23 


An  inconsistency  in  plate  series  definition 

The  foregoing  comments  are  specifically  concerned  with  the  anterior  intercalary  and  apical  series  in 
gonyaulacacean  epithecae.  More  important  is  the  concept  of  these  series  in  peridiniacean  dino- 
flagellates.  The  partially  developed  anterior  intercalary  series  is  a characteristic  feature  of  the 
peridiniacean  plate  pattern,  and  there  can  be  little  doubt  that  Kofoid  considered  these  intercalaries  to 
be  additional  plates  between  the  precingulars  and  apicals.  However,  a polar  view  of  the  peridiniacean 
epitheca  does  not  support  this  interpretation.  In  Protoperidinium  depressum  (text-fig.  2)  for  instance, 
the  three  anterior  intercalaries  la-3a  and  apicals  T,  2',  and  4'  form  a perfect  ring  of  plates,  effectively 
concentric  with  the  precingular  series.  The  interpretation  of  these  six  plates  as  the  apical  series  would 
leave  only  Kofoidian  apical  3'  unaccounted  for.  Thus  Kofoid’s  concept  of  the  anterior  intercalary 
series  seems  to  be  an  artificial  one  which  resulted  directly  from  his  interpretation  of  the  apical  series  in 
peridiniacean  dinoflagellates. 

I can  only  speculate  on  the  reasons  why  Kofoid  defined  the  apical  series  in  the  way  he  did.  He  may 
simply  have  believed  that  the  apical  plates  should  occupy  or  at  least  touch  the  morphological  apex. 
He  may  have  been  influenced  by  the  fact  that  his  concept  of  the  apical  series  resulted  in  the 
recognition  of  four  apical  plates  in  Ceratium,  peridiniacean  taxa,  and  certain  species  of  Gonyaulax, 
and  this  consistency  might  be  significant.  Whatever  reason  is  suggested,  one  major  criticism  is 
inescapable:  Kofoid’s  concept  of  the  apical  series  in  peridiniacean  dinoflagellates  is  incompatible 
with  his  basic  statement  on  plate  series  definition.  That  is,  since  the  division  of  the  theca  into  epitheca 
and  hypotheca  is  of  such  fundamental  importance,  the  intervening  cingulum  should  be  used  as  the 
basis  for  defining  the  transverse  plate  series  (Kofoid  1909,  pp.  41,  43). 

For  consistency,  after  the  precingular  series  had  been  defined  as  the  plates  anterior  to  and 
contiguous  to  the  cingulum,  the  next  series  should  have  been  defined  as  the  plates  anterior  to  and 
contiguous  to  the  precingulars.  This  consistent  definition  of  the  apical  series  would  not  have  affected 
Kofoid’s  interpretation  of  Ceratium,  but  it  would  have  greatly  affected  his  interpretation  of  the 
peridiniacean  plate  pattern.  In  P.  depressum  there  would  be  six  apicals  rather  than  four,  a residue  of 
one  plate  at  the  apex  (Kofoid’s  apical  3'),  and  no  anterior  intercalaries.  In  species  of  Gonyaulax  such 
as  G.  polyedra  there  would  be  five  apicals  (Kofoid’s  l'-3',  la,  2a)  rather  than  four,  a residue  of  two 
plates  at  the  apex  (1  ap.  cl.  and  Kofoid’s  apical  4'),  and  again  no  anterior  intercalaries.  In  H.  clado- 
phora  there  would  be  five  apicals,  in  Ctenidodinium  sp.  there  would  be  six,  and  the  conventional 
intercalary  plate  in  both  patterns  previously  suggested  to  be  homologous  with  apical  3'  in  G.  polyedra 
would  now  be  designated  apical.  Also,  apical  4'  in  C.  hirundinella  would  be  designated  apical, 
independent  of  its  relationship  with  the  morphological  apex. 

A similar  argument  can  be  made  against  Kofoid’s  interpretation  of  the  antapical  series  in 
Gonyaulax  and  his  resulting  concept  of  a posterior  intercalary  series.  After  the  postcingular  series  had 
been  defined  as  the  plates  posterior  to  and  contiguous  to  the  cingulum,  the  next  series  should  have 
been  defined  as  the  plates  posterior  to  and  contiguous  to  the  postcingulars.  This  consistent  definition 
of  the  antapical  series  would  not  have  affected  Kofoid’s  interpretation  of  Ceratium  or  the  basic 
peridiniacean  plate  pattern,  but  it  would  have  affected  his  interpretation  of  Gonyaulax.  In  the  latter 
genus,  the  plate  conventionally  designated  posterior  intercalary  1 p would  become  first  antapical  1 
conventional  \""  would  become  2"",  and  there  would  be  no  posterior  intercalaries. 

Thus  initially  on  the  grounds  of  consistency  in  plate  series  definition  and  some  limited  evidence  of 
plate  homology,  redefinition  of  the  apical  and  antapical  series  is  justified. 


MODIFICATION  OF  KOFOID’S  SYSTEM  OF  PLATE  SERIES  NOMENCLATURE 

Definition  of  all  the  transverse  plate  series  relative  to  the  cingulum  generally  results  in  the  recognition  of  two 
major  plate  series  on  both  the  epitheca  and  hypotheca.  Any  remaining  plates  occur  at  or  near  the  poles  of  the 
theca  and  can  be  accommodated  in  a third  epithecal  or  hypothecal  series.  Although  this  approach  differs  from 
Kofoid’s  concept  of  transverse  plate  series,  only  the  apical  and  antapical  series  need  to  be  redefined.  Also,  almost 
all  of  Kofoid’s  terms  are  still  applicable  and  there  is  only  one  completely  new  plate  series. 

I would  emphasize  here  that  the  following  definitions  are  only  intended  to  be  broad  guides  to  the  recognition 


EATON:  DINOFL AGELLATE  PLATE  PATTERNS 


673 


of  the  various  plate  series  and  the  designation  of  individual  plates.  I do  not  believe  that  such  definitions  should  be 
rigidly  applied.  Subjective  interpretation  is  unavoidable,  and  interplate  relationships  must  be  considered  for 
each  plate  pattern  before  individual  plates  can  be  assigned  to  the  various  plate  series. 

Definition  of  the  transverse  series 

Three  transverse  series  are  recognized  on  the  epitheca:  precingular,  apical,  apical  closing;  and  three  are  also 
recognized  on  the  hypotheca:  postcingular,  antapical,  antapical  closing. 

The  precingular  series  (")  was  satisfactorily  defined  by  Kofoid  (1907,  p.  179)  as  the  row  of  plates  anterior  to 
and  contiguous  to  the  cingulum.  This  definition  is  retained  here. 

The  apical  series  (')  is  redefined  as  the  row  of  plates  anterior  to  and  contiguous  to  the  precingular  series.  Also 
included  is  the  plate  (or  plates)  anterior  to  and  contiguous  to  the  sulcal  area,  as  suggested  by  Kofoid.  The  apical 
series  may  be  interrupted  by  a posterior  extension  of  the  apical  closing  series  and  in  certain  circumstances  apical 
plates  may  touch  the  cingulum  (e.g.  Helgolandinium  subglobosum,  text-fig.  7). 

The  apical  closing  series  (ap.  cl.)  is  defined  as  the  plates  anterior  to  and  contiguous  to  the  apical  series.  The 
concept  of  apical  closing  plates  was  discussed  by  Kofoid  (191 1,  p.  194)  with  respect  to  the  small  plate  occupying 
the  extreme  apex  of  Gonyaulax.  My  idea  of  the  apical  closing  series  includes  this  and  any  other  plates  anterior 
and  contiguous  to  the  apical  series,  with  the  term  ‘closing’  being  used  in  a geometric  rather  than  a biologically 
functional  sense.  This  series  may  be  represented  by  a distinct  row  of  plates,  and  in  certain  circumstances  apical 
closing  plates  may  interrupt  the  apical  series  and  touch  the  precingular  series  (e.g.  Shublikodinium  arcticum,  text- 
fig.  8). 

The  postcingular  series  ('")  was  satisfactorily  defined  by  Kofoid  (1907,  p.  179)  as  the  row  of  plates  posterior  to 
and  contiguous  to  the  cingulum.  This  definition  is  retained  here. 

The  antapical  series  ("")  is  redefined  as  the  plates  posterior  to  and  contiguous  to  the  postcingular  series. 

The  antapical  closing  series  (an.  cl.)  is  proposed  as  a new  series,  and  is  defined  as  the  plates  posterior  to  and 
contiguous  to  the  antapical  series.  Again,  ‘closing’  is  used  in  a purely  geometric  sense.  So  far  this  series  has  been 
recognized  only  in  S.  arcticum  and  Rhaetogonyaulax  rhaetica  (both  text-fig.  8). 

Application  to  selected  modern  and  fossil  dinoflagellate  plate  patterns 

The  plate  patterns  of  five  modern  and  ten  fossil  dinoflagellates  are  illustrated  in  text-figs.  4-8  as  diagrammatic 
polar  (epithecal  and  hypothecal)  views,  in  which  I have  tried  to  retain  true  interplate  relationships  with  minimum 
distortion  of  observed  plate  geometry.  This  type  of  illustration  is  used  rather  than  conventional  ventral  and 
dorsal  views  (text-fig.  1)  because  it  allows  a better  appreciation  of  the  geometric  relationship  between  individual 
plates  or  groups  of  plates.  Hypothecae  are  illustrated  with  the  sulcus  to  the  south  rather  than  the  conventional 
north,  to  emphasize  certain  similarities  with  epithecae.  The  plates  are  numbered  in  terms  of  the  modified  plate 
series  nomenclature,  and  the  direction  of  numbering  is  conventional.  Individual  cingular  and  sulcal  plates  are 
not  indicated. 

The  listed  data  for  each  pattern  include  the  specific  name  with  its  authorship,  geological  age  (where  relevant), 
the  source  of  the  plate  pattern,  and  the  modified  tabulation  formula.  The  formula  is  expressed  in  terms  of  the 
epithecal  (E)  and  hypothecal  (H)  transverse  series  only.  Changes  in  plate  designation  are  indicated,  with  the 
reinterpreted  designation  first,  followed  by  the  conventional  Kofoidian  designation  in  parentheses.  Where 
necessary,  changes  in  dinoflagellate  cyst  archaeopyle  nomenclature  are  indicated  for  the  fossil  taxa,  in  terms  of 
the  notation  previously  discussed  by  Evitt  (1967)  and  Stover  and  Evitt  (1978). 

Gonyaulax  spinifera  (Claparede  and  Lachmann)  Diesing  1 866,  Recent,  text-fig.  4. 

From:  Kofoid  (1911,  text-figs,  a-d)  and  Wall  and  Dale  (1970,  text-figs.  19-22). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  4',  6";  H:  6"',  2'"'. 

Changes  in  plate  designation:  1"”  (lp),  2""  (1"")- 

Gonyaulax  polyedra  Stein  1883,  Recent,  text-fig.  4. 

From:  Kofoid  (1911,  pi.  12,  figs.  16-20). 

Modified  tabulation  formula,  E:  2 ap.  cl.,  5',  6";  H:  6'",  2"" . 

Changes  in  plate  designation:  2 ap.  cl.  (4'),  4',  5'  (la,  2a),  V"  (lp),  2""  (1""). 

Hystrichogonyaulax  cladophora  (Deflandre)  Stover  and  Evitt  1978,  Late  Jurassic,  text-fig.  4. 

From:  Deflandre  (1938,  text-figs.  5,  6)  and  my  own  observations. 

Modified  tabulation  formula,  E:  1 ap.  cl.,  5',  6";  H:  6'",  2"" . 

Changes  in  plate  designation:  3'  (la),  4',  5',  (3',  4'),  1""  (lp),  2""  (1'"')- 


674 


PALAEONTOLOGY,  VOLUME  23 


Ctenidodinium  Deflandre  1938  sp.,  Middle  Jurassic,  text-fig.  5. 

From:  my  own  observations. 

Modified  tabulation  formula,  E:  2 ap.  cl.,  6',  6";  H:  6'",  2"" . 

Changes  in  plate  designation:  3',  4'  (la,  2a),  5',  6'  (3\  4'),  1""  (lp),  2.""  (1""). 

Paragonyaulacysta  Johnson  and  Hills  1973  s.l. , Middle  Jurassic,  text-fig.  5. 

From:  Johnson  and  Hills  (1973,  text-fig.  9)  and  my  own  observations. 

Modified  tabulation  formula,  E:  1 ap.  cl.,  5',  6";  H:  6"',  2"" . 

Changes  in  plate  designation:  3'-5'  (la-3a),  1""  (lp),  2"”  (1""). 

Cyst  archaeopyle  type:  dorsal  apical,  type  A,  2A  or  3A  (conventionally  intercalary,  type  I,  21  or  31). 
Luehndea  spinosa  Morgenroth  1970,  Early  Jurassic,  text-fig.  5. 

From:  Morgenroth  (1970,  pi.  9,  figs.  1-4),  Evitt  (unpublished  data). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  6',  6";  H:  6"',  2"" . 

Changes  in  plate  designation:  1 ap.  cl.  (3'),  3'-5'  (la-3a),  6'  (4'),  1""  (lp),  2""  ( 1 

Canninginopsis  denticulata  Cookson  and  Eisenack  1962,  Mid  Cretaceous,  text-fig.  6. 

From:  Cookson  and  Eisenack  (1962,  text-fig.  2),  Wall  and  Evitt  (1975,  text-fig.  11). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  4',  6";  H:  6'",  2”''. 

Changes  in  plate  designation:  1""  (lp),  2""  (1""). 


text-fig.  4.  Modified  interpretation  of  tabulation  in  polar  views  of  Gonyaulax  spinifera  (Claparede  and 
Lachmann)  Diesing  1866,  Gonyaulax  polyedra  Stein  1883,  Hystrichogonyaulax  cladophora  (Deflandre)  Stover 
and  Evitt  1978.  Upper,  epithecae.  Lower,  hypothecae. 


EATON:  DINOFLAGELL ATE  PLATE  PATTERNS 


675 


Ceratium  hirundinella  (Muller)  Schrank  1793,  Recent,  text-fig.  6. 

From:  Wall  and  Evitt  (1975,  text-figs.  5,  6). 

Modified  tabulation  formula,  E:  4',  6";  H:  6"',  2"". 

Changes  in  plate  designation:  1""  (lp),  2""  (1""),  relative  to  Wall  and  Evitt  (1975). 

Thalassiphora  delicata  Williams  and  Downie  1966,  Eocene,  text-fig.  6. 

From:  Eaton  (1976,  text-figs.  18,  20). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  4',  6";  H:  6"',  2'"'. 

Changes  in  plate  designation:  1 ap.  cl.  (4'),  4'  (la),  2'"'  (lp).  Also,  the  sixth  pre-  and  postcingulars  are  now 
recognized. 

Protoperidinium  depressum  (Bailey)  Balech  1974,  Recent,  text-fig.  7. 

From:  Gocht  and  Netzel  (1974,  text-fig.  1). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  6',  7";  H:  5'",  2"". 

Changes  in  plate  designation:  1 ap.  cl.  (3'),  3'-5'  (la-3a),  6'  (4'). 

Phthanoperidinium  tritonium  Eaton  1976,  Eocene,  text-fig.  7. 

From:  Eaton  (1976,  text-fig.  24). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  6',  7";  FI:  5'”,  2"". 

Changes  in  plate  designation:  1 ap.  cl.  (3'),  3'-5'  (la-3a),  6'  (4'). 

Cyst  archaeopyle  type:  dorsal  apical,  type  A (conventionally  intercalary,  type  I). 


text-fig.  5.  Modified  interpretation  of  tabulation  in  polar  views  of  Ctenidodinium  Deflandre  1938  sp., 
Paragonyaulacysta  Johnson  and  Hills  1973  s.l.,  Luehndea  spinosa  Morgenroth  1970.  Upper,  epithecae. 

Lower,  hypothecae. 


676 


PALAEONTOLOGY,  VOLUME  23 


Helgolandinium  subglobosum  von  Stosch  1969,  Recent,  text-fig.  7. 

From:  von  Stosch  (1969,  text-fig.  3). 

Modified  tabulation  formula,  E:  5',  7";  H:  1"',  3"”. 

Changes  in  plate  designation:  T (1"),  2'-5'  (l'-4'),  T'-7"  (2"-8"). 

The  very  small  plate  designated  9"  by  von  Stosch  (1969,  text-fig.  3/)  which  lies  anterior  to  the  anterior  sulcal 
plate  and  touches  the  first  cingular  plate  is  here  referred  to  the  sulcus  as  a second  anterior  sulcal  plate,  2 a.s.  A 
similarly  positioned  plate  can  sometimes  be  recognized  in  Paragonyaulacysta  s.l. 

Dapcodinium  priscum  Evitt  1961,  Early  Jurassic,  text-fig.  8. 

From:  Evitt  (1961,  text-figs.  1-20). 

Modified  tabulation  formula,  E:  1 ap.  cl.,  1',  7";  H:  7'",  3"". 

Changes  in  plate  designation:  1 ap.  cl.  (3'),  3'-6'  (la-4a),  7'  (4'),  2"'-7"'  (T"-6,,,)J  1"",  2""  (lp,  2p),  3""  (1""). 
On  the  hypo  theca,  Evitt  originally  recognized  only  six  postcingulars,  but  several  of  his  drawings  (Evitt  1961, 
text-figs.  5- 10, 1 5- 1 7, 1 9)  show  an  undesignated  plate  comparable  in  position  to  the  reduced  first  postcingular  of 
Gonyaulax.  This  plate  is  designated  T"  here,  and  the  number  of  postcingulars  is  increased  from  six  to  seven. 

Shublikodinium  arcticum  Wiggins  1973,  Late  Triassic,  text-fig.  8. 

From:  Wiggins  (1973,  text-fig.  3). 

Modified  tabulation  formula,  E:  6 ap.  cl.,  6',  7";  H:  7"',  3"",  1 an.  cl. 

Changes  in  plate  designation:  1 ap.  cl.  (va.  cl.),  2-6  ap.  cl.  ( 2'-6 '),  2'-6'  (la-5a),  1 an.  cl.  (ppl). 

Cyst  archaeopyle  type:  combination  apical  closing-apical,  type  tACL  tA  (conventionally  combination  apical- 
intercalary,  type  tA  tl). 


text-fig.  6.  Modified  interpretation  of  tabulation  in  polar  views  of  Canninginopsis  denticulata  Cookson  and 
Eisenack  1962,  Ceratium  hirundinella  ( Muller)  Schrank  1793,  Thalassiphora  delicata  Williams  and  Downie  1966. 
Upper,  epithecae.  Lower,  hypothecae. 


P.  DEPRESSUM 


text-fig.  7.  Modified  interpretation  of  tabulation  in  polar  views  of  Protoperidinium  depressum  (Bailey)  Balech 
1974,  Phthanoperidinium  tritonium  Eaton  1976,  Helgolandinium  subglobosum  von  Stosch  1969.  Upper,  epithecae. 

Lower,  hypothecae. 


text-fig.  8.  Modified  interpretation  of  tabulation  in  polar  views  of  Dapcodinium  priscum  Evitt  1961, 
Shublikodinium  arcticum  Wiggins  1973,  Rhaetogonyaulax  rhaetica  (Sarjeant)  Loeblich  and  Loeblich  1968. 
Upper,  epithecae.  Lower,  hypothecae. 


678 


PALAEONTOLOGY,  VOLUME  23 


Rhaetogonyaulax  rhaetica  (Sarjeant)  Loeblich  and  Loeblich  1968,  Late  Triassic,  text-fig.  8. 

From:  Harland,  Morbey  and  Sarjeant  (1975,  text-fig.  2). 

Modified  tabulation  formula,  E:  6 ap.  cl.,  7',  7";  H:  7'",  3"",  1 an.  cl. 

Changes  in  plate  designation:  1-6  ap.  cl.  (l'-6'),  1'  (a.v.),  2'-7'  (la-6a),  1"",  (lp),  1 an.  cl.  (1""). 

Cyst  archaeopyle  type:  combination  apical  closing-apical,  type  tACL  tA  (conventionally  combination  apical- 
intercalary,  type  tA  tl). 


HOMOLOGOUS  AND  CORRESPONDING  PLATES  IN  SELECTED 
DINOFLAGELLATE  PLATE  PATTERNS 

Method 

The  method  used  here  for  recognizing  homologous  and  corresponding  plates  in  different  plate 
patterns  involves  critical  comparisons  of  their  interplate  relationships.  These  comparisons  are  made 
with  respect  to  a model  plate  pattern  (text-fig.  9)  whose  epitheca  and  hypotheca  both  show  a very  high 
degree  of  radial  symmetry  and  plate  regularity.  The  model  plate  series  are  interpreted  in  terms  of  the 
modified  nomenclature,  but  to  emphasize  the  model  nature  of  the  pattern,  the  Kofoidian  style 
notation  is  not  applied.  Instead,  the  series  are  referred  to  as  ‘ap.’  (apical),  ‘prec.’  (precingular), 
‘postc.’  (postcingular)  and  ‘antap.’  (antapical).  Also,  the  plates  in  each  series  are  simply  numbered 
consecutively  1,  2,  3 etc.,  and  referred  to  as  ap.  1,  postc.  3,  etc.  The  idea  of  this  model  plate  pattern 
is  based  on  the  following  observations. 

In  the  fifteen  epithecal  patterns  illustrated  in  text-figs.  4-8,  the  maximum  number  of  plates  in  any  of 
the  transverse  series  is  seven  (7',  e.g.  D.  priscum\  7",  e.g.  P.  depressum).  Also,  the  most  regular 
interseries  relationship  involves  groups  of  three  plates  (3-plate  relationship)  with  one  plate  in  one 
series  touching  two  plates  in  an  adjacent  series.  This  relationship  is  fully  developed  between  the 
precingular  and  apical  series  in  L.  spinosa  and  D.  priscum  for  instance,  and  between  six  of  the  apicals 
(2'-T)  and  five  of  the  apical  closing  plates  (2-6  ap.  cl.)  in  R.  rhaetica. 

The  model  epitheca  shows  a full  development  of  the  3-plate  interseries  relationship  in  a pattern 
with  seven  plates  in  all  three  transverse  series.  The  direction  of  numbering  the  epithecal  plates  is 
conventional,  i.e.  anticlockwise  relative  to  the  apical  pole.  When  an  observed  epithecal  pattern  does 
not  show  counterparts  of  all  the  model  plates,  it  is  the  highest  numbered  model  plates  which  are 
considered  to  be  unrepresented. 

A similar  interpretation  of  such  observations  on  the  fifteen  hypothecal  patterns  illustrated  in  text- 
figs.  4-8  would  lead  to  a model  pattern  closely  comparable  to  S.  arcticum  and  R.  rhaetica.  The 


text-fig.  9.  The  model  plate  pattern.  Left,  epitheca,  E:  7 ap.  cl.,  7 ap.,  7 prec.  Right, 
hypotheca,  H:  7 postc.,  7 antap.,  1 an.  cl. 


EATON:  DINOFLAGELLATE  PLATE  PATTERNS 


679 


maximum  number  of  plates  in  any  of  the  transverse  series  is  again  seven,  but  this  maximum  only 
occurs  in  the  postcingular  series  (e.g.  H.  subglobosum).  The  maximum  number  of  antapicals  is  only 
three  (e.g.  D.  priscum),  with  a single  antapical  closing  plate  developed  only  in  S.  arcticum  and 
R.  rhaetica.  All  four  patterns  with  seven  postcingulars  and  three  antapicals  ( H . subglobosum, 
D.  priscum,  S.  arcticum,  R.  rhaetica)  show  a constant  relationship  between  these  two  series.  This 
involves  groups  of  four  plates  (4-plate  relationship)  with  one  antapical  touching  three  postcingulars. 
However,  this  arrangement  can  be  interpreted  in  terms  of  the  fundamental  3-plate  relationship  (fully 
developed  in  the  model  epitheca),  if  each  of  the  three  antapicals  is  treated  as  two  plates,  and  the 
intervening  mid-ventral  area  is  also  considered  to  be  an  antapical  plate. 

Thus  the  model  hypotheca  shows  a full  development  of  the  3-plate  relationship  between  the 
postcingulars  and  antapicals,  with  seven  plates  in  both  series.  The  model  hypothecal  plates  are 
numbered  anticlockwise  relative  to  the  antapical  pole.  This  is  the  reverse  of  convention,  but  is  more 
convenient  for  the  discussion  of  observed  hypothecal  patterns  which  do  not  show  counterparts  of  all 
the  model  plates.  As  with  the  epithecae,  it  is  the  highest  numbered  model  plates  which  are  considered 
to  be  unrepresented. 

In  text-fig.  9,  four  plates  on  both  the  epitheca  and  hypotheca  are  ornamented.  I consider  these  to  be 
key  reference  plates  in  the  discussion  of  homologous  plate  relationships  and  the  comparison  of 
different  plate  patterns. 

For  ease  of  comparison  with  the  model  pattern,  the  epithecae  and  hypothecae  of  the  selected 
modern  and  fossil  dinoflagellates  are  discussed  and  illustrated  separately  (text-figs.  10-12).  The 
modified  system  of  plate  series  nomenclature  is  applied  throughout  with  a Kofoidian  style  notation. 


text-fig.  10.  Epithecal  plate  patterns  of  six  selected  dinoflagellate  taxa.  The  key  reference  areas  are  ornamented. 


680 


PALAEONTOLOGY,  VOLUME  23 


The  direction  of  numbering  the  plates  is  conventional.  In  the  text-figures  the  interseries  boundaries 
are  thickened  for  emphasis,  and  only  those  plates  critical  to  the  discussion  are  numbered.  A 
maximum  of  four  such  plates  on  both  the  epitheca  and  hypotheca  are  also  ornamented,  either  in  full 
or  in  part.  Each  ornamented  area  corresponds  to  one  key  plate  in  the  model  pattern. 

Epithecal  plate  patterns  (text-figs.  9,  10,  11) 

The  model  epithecal  pattern  (text-fig.  9)  has  twenty-one  plates  arranged  in  three  series  (7  ap.  cl.,  7 ap., 
7 prec.).  This  discussion  is  primarily  concerned  with  the  apical  and  precingular  series  in  which  the  key 
reference  plates  are  ap.  1,  ap.  4 (stippled),  and  prec.  4,  prec.  6 (shaded). 

Rhaetogonyaulax  rhaetica  and  Dapcodinium  priscum  (text-fig.  10)  both  have  a 7',  7"  pattern,  and 
also  show  a full  development  of  the  3-plate  relationship.  Because  of  this,  the  seven  apicals  and  seven 
precingulars  in  both  patterns  are  considered  to  be  respectively  homologous  with  ap.  1-7  and 
prec.  1-7. 

In  Shublikodinium  arcticum  (text-fig.  10)  which  has  a 6'  1"  pattern,  the  3-plate  relationship  is  lost  in 
the  vicinity  of  precingular  2"  which  is  touched  by  only  one  apical,  2'.  Compared  with  the  model 
pattern,  plate  2'  occupies  the  position  of  two  plates,  ap.  2 and  ap.  3.  Thus  apical  T in  S.  arcticum 
corresponds  to  two  plates  in  the  model  pattern,  and  as  a result,  apical  3'  in  S.  arcticum  is  homologous 
with  key  ap.  4.  The  interruption  of  the  apical  series  in  S.  arcticum  by  a posterior  extension  of  the 
apical  closing  series  between  T and  2',  does  not  affect  the  over-all  interpretation  of  the  3-plate 
relationship.  Phthanoperidinium  tritonium  (text-fig.  10)  also  has  a 6',  7"  pattern,  but  the  3-plate 
relationship  is  lost  in  the  vicinity  of  4”.  Thus  apical  4'  corresponds  to  two  plates  in  the  model  pattern, 
and  only  its  stippled  area  corresponds  to  key  ap.  4.  In  the  6',  1"  pattern  of  Protoperidinium  depressum 
(text-fig.  10),  the  3-plate  relationship  is  lost  in  the  vicinity  of  three  consecutive  precingulars,  3" -5”. 
Even  so,  I still  consider  that  only  apical  4'  corresponds  to  two  plates  in  the  model  pattern,  and  this 
accounts  for  the  loss  of  the  3-plate  relationship  between  the  apical  series  and  4”.  The  further  loss  in 
the  vicinity  of  3"  and  5"  is  due  to  a relative  lateral  displacement  of  the  two  intra-apical  sutures  which 
border  4'.  Thus  only  the  stippled  area  of  4'  in  P.  depressum  corresponds  to  key  ap.  4.  In 
Helgolandinium  subglobosum  (text-fig.  10)  which  has  a 5',  7"  pattern,  the  3-plate  relationship  is  lost  in 
the  vicinity  of  two  separated  precingulars,  2"  and  5".  Thus  apicals  2'  and  4'  each  correspond  to  two 
plates  in  the  model  pattern,  and  3'  is  homologous  with  key  ap.  4. 

In  Luehndea  spinosa  (text-fig.  11)  there  are  only  six  plates  in  both  the  apical  and  precingular  series 
(6',  6"  pattern),  but  the  3-plate  relationship  is  maintained  throughout.  Because  of  this  the  twelve 
plates  comprising  the  apical  and  precingular  series  are  considered  to  be  respectively  homologous  with 
ap.  1-6  and  prec.  1-6.  Thus  the  highest  numbered  model  plates,  ap.  7 and  prec.  7,  have  no 
counterparts  in  L.  spinosa.  This  last  comment  also  applies  to  the  eight  remaining  patterns  discussed 
here. 

Ctenidodinium  sp.  (text-fig.  1 1)  has  a 6',  6"  pattern,  but  the  3-plate  relationship  is  lost  in  the  vicinity 
of  two  consecutive  precingulars,  2"  and  3".  Plate  2"  has  one  apical  touching  it,  2',  while  3"  has  three 
apicals  touching  it,  2,-4/.  The  loss  of  the  3-plate  relationship  is  due  to  the  critical  shortening  of  3', 
accompanied  by  the  relative  enlargement  of  2'.  Plate  4'  is  also  shortened,  but  not  critically.  As  in 
L.  spinosa , the  twelve  plates  comprising  the  apical  and  precingular  series  in  Ctenidodinium  sp.  are 
respectively  homologous  with  ap.  1-6  and  prec.  1-6. 

In  Gonyaulax  polyedra  and  Hystrichogonyaulax  cladophora  (text-fig.  11)  which  both  have  a 5',  6” 
pattern,  the  3-plate  relationship  is  lost  in  the  vicinity  of  precingular  2".  Applying  the  principle  used  in 
the  interpretation  of  S.  arcticum,  P.  tritonium  etc.,  apical  2'  corresponds  to  two  plates  in  the  model 
pattern.  Thus  3'  in  G.  polyedra  and  H.  cladophora  is  homologous  with  key  ap.  4.  The  shortening  of  3' 
in  H.  cladophora  is  not  critical,  and  is  comparable  to  4'  in  Ctenidodinium  sp.  In  Par  agony  aulacysta  s.l. 
(text-fig.  11)  which  also  has  a 5',  6"  pattern,  the  3-plate  relationship  is  lost  in  the  vicinity  of 
precingular  6".  Thus  only  the  stippled  area  of  1 ' corresponds  to  key  ap.  1 , and  the  unornamented  area 
of  T corresponds  to  ap.  6.  The  shortening  of  apicals  3' -5'  in  Par  agony  aulacysta  s.l.  is  not  critical. 

Gonyaulax  spinifera,  Canninginopsis  denticulata,  and  Ceratium  hirundinella  (text-fig.  1 1)  all  have  a 
4\  6”  pattern  and  identical  homologous  and  corresponding  plate  relationships.  Loss  of  the  3-plate 


EATON:  DINOFLAGELLATE  PLATE  PATTERNS 


L.  SPINOSA  CTENIDODINIUM  S P.  G.  POLYEDRA 


H.  CLADOPHORA  PA  R A G O N Y A U L A C Y S T A S.L.  G.  SPINIFERA 


C.  DENTICULATA 


C.  HIRUNDINELLA 


T.  DELICATA 


text-fig.  1 1 . Epithecal  plate  patterns  of  a further  nine  selected  dinoflagellate  taxa.  The  key  reference  areas  are 

ornamented. 


682 


PALAEONTOLOGY,  VOLUME  23 


relationship  occurs  in  the  vicinity  of  precingulars  2"  and  4".  Thus  2'  and  3'  each  correspond  to  two 
plates  in  the  model  pattern,  and  therefore  the  stippled  area  of  3'  corresponds  to  key  ap.  4.  The  further 
loss  of  the  3-plate  relationship  in  G.  spinifera  is  due  to  the  critical  shortening  of  6".  The  shortening  of 
4'  in  C.  hirundinella  is  not  critical. 

In  Thalassiphora  delicata  (text-fig.  1 1)  which  also  has  a 4',  6"  pattern,  the  3-plate  relationship  is  lost 
in  the  vicinity  of  three  precingulars,  2",  5",  and  6".  Thus  2'  and  4'  each  correspond  to  two  plates  in  the 
model  pattern,  and  3'  is  homologous  with  key  ap.  4.  Plate  6”  is  critically  shortened  and  is  not  touched 
by  any  of  the  apical  series. 

Only  limited  comments  can  be  made  on  possible  homologous  relationships  in  the  apical  closing 
series,  because  of  the  great  variation  in  its  development.  It  is  reasonable  to  suggest  that  1 ap.  cl.  in 
R.  rhaetica  and  S.  arcticum  corresponds  to  the  first  and  seventh  apical  closing  plates  in  the  model 
pattern,  and  that  the  five  remaining  plates  in  all  three  patterns  are  respectively  homologous.  Also  that 
the  large  single  apical  closing  plate  in  D.  priscum,  L.  spinosa,  P.  tritonium,  and  P.  depressum 
corresponds  to  all  seven  apical  closing  plates  in  the  model  pattern.  However,  the  relationship  between 
these  two  extremes  of  development,  and  the  apical  closing  plates  in  G.  spinifera  and  Ctenidodinium  sp. 
for  instance,  is  undetermined. 

Hypothecal  plate  patterns  (text-figs.  9,  12) 

The  model  hypothecal  pattern  (text-fig.  9)  has  fifteen  plates  arranged  in  three  series  (7  postc.,  7 antap., 
1 an.  cl.).  This  discussion  is  only  concerned  with  the  postcingular  and  antapical  series  in  which  the  key 
reference  plates  are  postc.  4,  postc.  6 (shaded)  and  antap.  3,  antap.  6 (stippled).  The  fifteen  selected 
hypothecae  do  not  show  the  same  range  of  variation  as  their  epithecae,  and  can  be  discussed  in  terms 
of  only  eight  patterns. 

The  ‘rhaetogonyaulacacean’  type  ( Rhaetogonyaulax  rhaetica,  Shublikodinium  arcticum),  Helgo- 
landinium  subglobosum,  and  Dapcodinium  priscum  (text-fig.  12)  all  have  a 1"' , 3""  pattern.  The  seven 
postcingulars  in  these  patterns  and  the  model  pattern  are  respectively  homologous.  The  antapical- 
postcingular  interseries  relationship  is  constant,  with  each  antapical  touching  three  postcingulars.  As 
stated  earlier,  each  antapical  corresponds  to  two  plates  in  the  model  pattern.  Since  occupy 

the  position  of  antap.  2-1,  corresponds  to  antap.  6-7,  2""  to  antap.  4-5,  and  3""  to  antap.  2-3. 
Thus,  only  the  stippled  area  of  3""  corresponds  to  key  antap.  3,  and  only  the  stippled  area  of  Y'" 
corresponds  to  key  antap.  6. 

Although  no  counterpart  of  antap.  1 is  shown  in  the  1'",  3"”  patterns,  it  may  be  represented  by  the 
posterior  area  of  the  sulcus.  However,  in  fossil  dinoflagellates  the  individual  sulcal  plates  are  often 
very  poorly  defined,  and  it  is  by  no  means  certain  that  all  areas  designated  posterior  sulcal  (p.s.)  are 
respectively  homologous.  Consequently  this  relationship  is  only  questionably  applied  to  the  fifteen 
selected  patterns  (Table  1). 

A significant  feature  of  D.  priscum  is  the  anticlockwise  rotation  of  its  plate  pattern  relative  to  the 
model  pattern,  and  the  associated  reduction  of  2"'  and  Y".  These  reduced  plates  correspond  to  the 
highest  numbered  model  postcingulars,  postc.  6,  postc.  7,  and  this  influenced  my  earlier  statement 
that  when  a hypothecal  pattern  does  not  show  counterparts  of  all  the  model  plates,  it  is  the  highest 
numbered  model  plates  which  are  unrepresented.  Consequently  it  is  convenient  to  discuss  the  plate 
relationships  of  the  five  remaining  patterns  in  an  anticlockwise  direction,  i.e.  in  terms  of  6' "-Y" 
and  2""-Y"'. 

In  Luehndea  spinosa  there  are  only  six  postcingulars  and  two  antapicals  ( 6 2""  pattern).  Plates 
6”'-Y"  are  respectively  homologous  with  postc.  1 -6,  and  there  is  no  counterpart  of  postc.  7.  Thus  3"' 
is  homologous  with  key  postc.  4,  and  1 is  homologous  with  key  postc.  6.  Antapical  2'"'  touches  four 
postcingulars  and  therefore  corresponds  with  three  plates  in  the  model  pattern,  antap.  2-4.  Thus  only 
the  stippled  area  of  2'"'  corresponds  to  key  antap.  3.  Plate  l""  touches  three  postcingulars  and 
therefore  corresponds  to  two  plates  in  the  model  pattern,  antap.  5,  6.  Thus  only  the  stippled  area  of 
Y'"  corresponds  to  key  antap.  6,  and  there  is  no  counterpart  of  antap.  7. 

In  the  ‘peridiniacean’  type  pattern  ( Protoperidinium  depressum  and  Phthanoperidinium  tritonium ) 
there  are  only  five  postcingulars  and  two  antapicals  (5'",  2""  pattern).  Plates  5"'-r"  are  respectively 


EATON:  DINOFL AGELL ATE  PLATE  PATTERNS 


683 


TY  PE  H.  SUBGLOBOSUM 


D.  PRISCUM 


PERIDINIACEAN 
L.  SPINOSA  TYPE 


TYPE 


C.  HIRUNDIN  ELLA 


DELICATA 


text-fig.  12.  Eight  hypothecal  plate  patterns  representative  of  the  fifteen  selected  dinoflagellate  taxa.  The  key 
reference  areas  are  ornamented. 


684 


PALAEONTOLOGY,  VOLUME  23 


table  1.  Homologous  and  corresponding  plates  in  the  fifteen  selected 
plate  patterns  and  the  model  pattern.  The  modified  interpretation  of 
transverse  series  is  applied  throughout.  Plates  with  an  asterisk  would  be 
designated  intercalary  using  the  conventional  Kofoidian  system. 


EATON:  DINOFLAGELLATE  PLATE  PATTERNS 


685 


homologous  with  postc.  1 -5  and  there  are  no  counterparts  of  postc.  6, 7.  Thus  2"'  is  homologous  with 
key  postc.  4.  Antapicals  2""  and  V"'  each  touch  three  postcingulars  and  therefore  each  corresponds 
to  two  plates  in  the  model  pattern.  Thus  only  the  stippled  area  of  2""  corresponds  to  key  antap.  3,  and 
there  are  no  counterparts  of  antap.  6,  7. 

The  ‘gonyaulacacean’  type  pattern  ( Gonyaulax  spinifera,  G.  polyedra,  Hystrichogonyaulax 
cladophora,  Ctenidodinium  sp.,  Canninginopsis  denticulata,  Paragonyaulacysta  s.l.)  with  six  post- 
cingulars and  two  antapicals  (6"',  2""  pattern)  differs  from  L.  spinosa  only  in  showing  considerable 
reduction  of  1 2"',  and  1 The  interseries  relationships  of  the  gonyaulacacean  pattern  are  directly 
comparable  with  L.  spinosa.  Thus  3'"  is  homologous  with  key  postc.  4,  and  Y"  is  homologous  with 
key  postc.  6.  Also,  only  the  stippled  area  of  2""  corresponds  to  key  antap.  3 and  only  the  stippled  area 
of  Y"'  corresponds  to  key  antap.  6. 

Ceratium  hirundinella  also  has  a 6'",  2""  pattern,  but  is  unusual  in  that  postcingular  6"'  touches 
V"  as  well  as  2"" . A possible  explanation  of  this  relationship  is  that  6'"  in  C.  hirundinella  actually 
represents  two  plates,  6'"  s.s.  and  the  conventional  posterior  sulcal  plate  (p.s.  in  text-fig.  1).  The  latter 
plate  has  not  been  precisely  identified  in  C.  hirundinella , although  it  must  be  admitted  that  the  sulcal 
area  of  Ceratium  in  general  is  poorly  known  (Wall  and  Evitt  1975,  p.  19).  Plates  5 ”'-Y"  are 
respectively  homologous  with  postc.  2-6,  and  there  is  no  counterpart  of  postc.  7.  Thus  3"'  is 
homologous  with  key  postc.  4,  and  1 is  homologous  with  key  postc.  6.  Antapical  2''"  touches  five 
postcingulars  and  therefore  corresponds  to  four  plates  in  the  model  pattern,  antap.  2-5.  Thus  only 
the  stippled  area  of  2""  corresponds  to  key  antap.  3.  Plate  \""  touches  V"  and  2"'  and  is  therefore 
homologous  with  key  antap.  6.  There  is  no  counterpart  of  antap.  7. 

Thalassiphora  delicata  also  has  a 6"',  2""  pattern,  but  its  interseries  relationships  effectively 
represent  a lateral  reversal  of  the  gonyaulacacean  arrangement.  This  affects  the  antapicals,  where  2”" 
only  touches  three  postcingulars  and  therefore  corresponds  to  two  plates  in  the  model  pattern,  and 
V"  touches  four  postcingulars  and  therefore  corresponds  to  three  plates  in  the  model  pattern.  Thus 
only  the  stippled  area  of  2""  corresponds  to  key  antap.  3,  and  only  the  stippled  area  of  1"" 
corresponds  to  key  antap.  6. 

The  interpreted  homologous  and  corresponding  plate  relationships  of  the  fifteen  selected  plate 
patterns  and  the  model  pattern  (excluding  the  apical  closing,  cingular  and  antapical  closing  series)  are 
summarized  in  Table  1 . This  emphasizes  the  fact  that  plates  conventionally  designated  intercalary  (*) 
in  one  pattern,  are  homologous  with  or  partially  correspond  to  conventional  apical  or  antapical 
plates  in  another  pattern.  Examples  include  apical  4'  (conventional  2a)  in  Protoperidinium  depressum 
homologous  with  apical  3'  in  Gonyaulax  spinifera , and  antapical  1 ""  (conventional  1 p)  in  G.  spinifera 
which  partially  corresponds  to  part  of  antapical  V”  in  Helgolandinium  subglobosum.  This  table  also 
shows  that  when  a series  is  represented  by  the  same  number  of  plates  in  different  patterns,  the  plates 
need  not  all  be  respectively  homologous.  Examples  include  the  four  apicals  in  G.  spinifera  and 
T.  delicata,  and  the  two  antapicals  in  G.  spinifera,  T.  delicata,  and  P.  depressum. 


DISCUSSION 

The  differences  between  the  fifteen  selected  plate  patterns  (text-figs.  10-12)  reflect  an  over-all  trend  of 
reduction  in  the  total  number  of  plates.  This  reduction  is  effected  in  two  particular  ways.  There  may 
be  simplification  through  the  development  of  a single  plate  in  one  pattern  which  spatially  corresponds 
with  two  or  more  plates  in  another  pattern.  This  critically  affects  interseries  relationships. 
Alternatively  there  may  be  a primary  development  of  fewer  plates  in  particular  series,  without 
affecting  interseries  relationships.  These  two  styles  of  reduction  may  occur  independently  or  together. 
They  may  also  be  accompanied  by  variation  in  the  relative  size  of  certain  plates  which  may  affect 
interseries  relationships  through  critical  shortening  or  lateral  reduction. 

Reduction  through  simplification  affects  the  apical  closing,  apical  and  antapical  series.  The 
development  of  a single  large  apical  closing  plate  in  D.  priscum  for  instance,  represents  simplification 
of  the  six-plate  arrangement  in  R.  rhaetica  and  S.  arcticum.  In  the  apical  series,  simplification  occurs 


686 


PALAEONTOLOGY,  VOLUME  23 


in  specific  areas,  e.g.  mid-ventral  (L)  in  Par  agony  aulacysta  s.l.;  left  lateral  (2')  in  S.  arcticum;  mid- 
dorsal (4')  in  P.  tritonium;  left  and  right  lateral,  (2',  3')  in  G.  spinifera,  (T,  4')  in  T.  delicata;  left  ventral 
and  right  lateral  (2',  4 ')  in  H.  subglobosum.  In  the  antapical  series,  simplification  is  best  defined  with 
reference  to  the  model  antapicals,  antap.  2-7.  For  instance,  the  three  antapicals  in  H.  subglobosum 
reflect  simplification  of  antap.  2-7  in  the  form  2-3, 4-5,  6-7.  In  the  patterns  with  only  two  antapicals 
this  simplification  takes  several  forms,  e.g.  2-3,  4-5  (peridiniacean  type),  2-4,  5-6  (gonyaulacacean 
type),  2-5,  6 (C.  hirundinella),  2-3,  4-6  (T.  delicata ). 

Reduction  through  the  primary  development  of  fewer  plates  is  best  defined  with  reference  to  the 
position  of  plates  which  are  homologous  with  or  in  part  correspond  to  model  key  reference  plates. 

In  the  six  epithecae  in  text-fig.  10  ( R . rhaetica  etc.)  the  position  of  these  plates  is  virtually  constant. 
In  particular,  the  equivalent  of  key  prec.  4.  is  invariably  mid-dorsal.  A similar  constancy  is  shown  by 
eight  of  the  epithecae  in  text-fig.  1 1 ( L . spinosa  etc.,  but  not  T.  delicata).  However,  in  these  patterns 
the  equivalent  of  key  prec.  4 is  invariably  right  dorso-lateral  in  position.  Compared  with  R.  rhaetica 
etc.  (text-fig.  10),  this  represents  a rotation  of  the  epithecal  pattern,  anticlockwise  relative  to  the 
apical  pole.  This  accommodates  the  primary  development  of  one  less  apical  and  one  less  precingular 
plate  (i.e.  no  counterparts  of  ap.  7,  prec.  7). 

In  the  eight  hypothecae  in  text-fig.  12  there  is  considerable  variation  in  the  position  of  the  key 
reference  areas.  In  particular,  the  equivalent  of  key  postc.  4 rotates  from  mid-dorsal  in  the 
rhaetogonyaulacacean  type  and  H.  subglobosum,  through  left  dorso-lateral  in  D.  priscum  and 
L.  spinosa,  to  left  lateral  in  the  gonyaulacacean  and  peridiniacean  types.  This  rotation,  which  affects 
the  over-all  hypothecal  pattern,  is  anticlockwise  relative  to  the  antapical  pole.  This  accommodates 
the  primary  development  of  fewer  postcingulars  and  antapicals  (i.e.  no  counterparts  of  two  or  more 
of  postc.  6,  7,  antap.  6,  7). 

The  effect  of  variation  in  the  relative  size  of  certain  plates  is  well  shown  by  the  epithecae  of 
Ctenidodinium  sp.,  H.  cladophora.  Par  agony  aulacysta  s.l.  and  G.  spinifera  (text-fig.  1 1),  in  which  there 
is  enlargement  of  the  lateral  and  left  ventral  precingulars  (compared  with  L.  spinosa ).  This  is  at  the 
expense  of  the  apical  series  which  becomes  longitudinally  aligned,  and  6”  which  is  reduced.  In  the 
somewhat  bizarre  pattern  of  T.  delicata  (text-figs.  11, 12)  the  considerable  enlargement  of  1 " and  2"', 
3"'  is  accommodated  by  the  displacement  of  the  sulcus,  1 ',  4',  1 6"'  and  \"" , 2""  to  a right  lateral 
position.  There  is  also  reduction  of  5"  and  6"',  and  6 " is  critically  shortened.  Other  examples  of 
critical  shortening  include  3'  in  Ctenidodinium  sp.  and  6"  in  G.  spinifera,  while  4'  in  P.  depressum  is 
critically  reduced  laterally. 

Reduction  in  the  total  number  of  thecal  plates  may  well  represent  a fundamental  trend  in  the 
evolution  of  peridinialean  plate  patterns.  If  this  is  so,  then  available  evidence  from  the  fossil  record 
suggests  that  primary  development  of  fewer  plates  was  the  most  important  means  of  achieving  this 
reduction.  This  evidence  is  provided  for  epithecae  by  the  appearance  in  the  Late  Triassic  of  patterns 
with  counterparts  of  ap.  7,  prec.  7 ( R . rhaetica,  S.  arcticum),  followed  in  the  Jurassic  by  the 
appearance  of  patterns  without  counterparts  of  these  two  plates  (e.g.  L.  spinosa,  Ctenidodinium  sp.). 
The  hypothecal  evidence  is  provided  by  the  successive  appearance  of  the  rhaetogonyaulacacean  type 
(Late  Triassic),  D.  priscum  and  L.  spinosa  (Early  Jurassic),  gonyaulacacean  type  (Middle  Jurassic) 
and  the  peridiniacean  type  (Late  Jurassic).  The  great  range  of  variation  shown  by  Late  Triassic  and 
younger  plate  patterns  resulted  from  the  effects  of  reduction  through  simplification,  and  variation  in 
relative  plate  size,  being  superimposed  on  the  effect  of  primary  reduction.  The  interaction  of  these 
three  variables  resulted  in  epithecae  and  hypothecae  evolving  comparatively  independently  and  this 
is  emphasized  by  the  way  in  which  the  epitheca  and  hypotheca  accommodated  the  effects  of  primary 
reduction.  In  both,  rotation  is  anticlockwise  relative  to  their  respective  pole,  and  therefore  the 
hypotheca  rotates  in  the  opposite  direction  to  the  epitheca  relative  to  the  polar  axis  of  the  theca.  The 
model  plate  pattern  appears  to  represent  an  evolutionary  base  to  which  all  Late  Triassic  and  younger 
peridinialeans  of  the  selected  type  are  related.  This  type  is  characterized  by  having  up  to  seven  plates 
in  each  of  its  epithecal  and  hypothecal  transverse  series.  In  view  of  this  relationship,  the  complex 
model  pattern  could  be  representative  of  a pre-Late  Triassic  ancestral  peridinialean. 


EATON:  DINOFL AGELL ATE  PLATE  PATTERNS 


687 


Acknowledgements.  I wish  to  acknowledge  the  considerable  advice  given  by  Barrie  Dale  (Oslo)  and  his  critical 
reading  of  the  manuscript.  The  ideas  expressed  in  this  paper  originally  formed  the  basis  of  a contribution  to  the 
GSA  Penrose  Conference  on  ‘Modern  and  Fossil  Dinoflagellates’,  held  in  Colorado  Springs,  Colorado,  U.S.A., 
April  1978. 1 wish  to  thank  the  many  workers  at  that  conference  who  provided  me  with  helpful  comments,  and 
particularly  Professor  W.  R.  Evitt  (Stanford)  who  also  critically  read  the  manuscript.  I am  grateful  to  the 
chairman  and  directors  of  British  Petroleum  Company  Limited  for  permission  to  publish,  to  Miss  Dorothy 
Watson  for  typing  the  manuscript,  and  to  my  wife  for  her  encouragement  during  the  preparation  of  this  paper. 


REFERENCES 

balech,  e.  1974.  El  genero  Protoperidinium  Bergh,  1881  ( Peridinium  Ehrenberg,  1831,  partim).  Rev.  Mus.  Cienc. 
natur.,  Hidrobiol.  4,  1-79. 

butschli,  o.  1885.  Unterabtheilung  (Ordnung)  Dinoflagellata.  Pp.  906-1029.  In  bronn,  h.  g.  Klassen  und 
Ordnung  des  Thier-Reichs,  wissenschaftlichen  dargestellt  in  Wort  und  Bild,  Band  1 Protozoa,  Abt.  2 Mastigo- 
phora.  Leipzig  and  Heidelberg. 

cookson,  i.  c.  and  eisenack,  a.  1962.  Additional  microplankton  from  Australian  Cretaceous  sediments.  Micro- 
paleontology, 8,  485-507,  pis.  1-7. 

deflandre,  G.  1938.  Microplancton  des  mers  Jurassiques  conserve  dans  les  marnes  de  Villers-sur-Mer 
(Calvados).  Etude  luminaire  et  considerations  generates.  Trav.  Stnzool.  Wimereux,  13,  147-200,  pis.  5-11. 
diesing,  K.  M.  1866.  Revision  der  Prothelminthen,  Abtheilung:  Mastigophoren.  Sber.  Akad.  Wiss.  Wien,  52, 
287-401. 

eaton,  g.  l.  1976.  Dinoflagellate  cysts  from  the  Bracklesham  Beds  (Eocene)  of  the  Isle  of  Wight,  southern 
England.  Bull.  Br.  Mus.  nat.  Hist.  (Geol.),  26,  225-332,  pis.  1-21. 
evitt,  w.  R.  1961.  Dapcodinium  priscum  n.gen.,  n.sp.,  a dinoflagellate  from  the  Lower  Lias  of  Denmark.  J. 
Paleont.  35,  996-1002,  pi.  119. 

— 1967.  Dinoflagellate  studies  II.  The  archeopyle.  Stanf.  Univ.  Pubis  (Geol.  Sci.),  10  (3),  1-88,  pis.  1-11. 

— lentin,  J.  K.,  millioud,  M.  E.,  stover,  l.  e.,  and  williams,  G.  L.  1976.  Dinoflagellate  cyst  terminology.  Geol. 
Surv.  Pap.  Can.  76-24,  1-9. 

GOCHT,  h.  1970.  Dinoflagellaten-Zysten  aus  dem  Bathonium  des  Erdolfeldes  Aldorf  (NW-Deutschland). 
Palaeontographica,  B 129,  125-165,  pis.  26-35. 

— and  netzel,  h.  1974.  Rasterelektronenmikroskopische  Untersuchungen  am  Panzer  von  Peridinium  (Dino- 
flagellata). Arch.  Protistenk.  116,  381-410,  pis.  43-52. 

harland,  r.,  morbey,  s.  J.,  and  sarjeant,  w.  a.  s.  1975.  A revision  of  the  Triassic  to  lowest  Jurassic  dinoflagellate 
Rhaetogonyaulax.  Palaeontology,  18,  847-864,  pis.  100-104. 

JOHNSON,  c.  D.  and  hills,  l.  v.  1973.  Microplankton  zones  of  the  Savik  Formation  (Jurassic),  Axel  Heiberg  and 
Ellesmere  Islands,  District  of  Franklin.  Bull.  Can.  petrol.  Geol.  21,  178-218,  pis.  1-3. 
kofoid,  c.  A.  1906.  Dinoflagellata  of  the  San  Diego  region.  I.  On  Heterodinium,  a new  genus  of  the  Peridinidae. 
Univ.  Calif.  Pubis  Zool.  2,  341-368,  pis.  17-19. 

— 1907.  The  plates  of  Ceratium  with  a note  on  the  unity  of  the  genus.  Zool.  Anz.  32,  177-183. 

— 1909.  On  Peridinium  steini  Jorgensen,  with  a note  on  the  nomenclature  of  the  skeleton  of  the  Peridinidae. 
Arch.  Protistenk.  16,  25-47,  pi.  2. 

— 1911.  Dinoflagellata  of  the  San  Diego  region,  IV.  The  genus  Gonyaulax , with  notes  on  its  skeletal 
morphology  and  a discussion  of  its  generic  and  specific  characters.  Univ.  Calif.  Pubis  Zool.  8,  187-286, 
pis.  9-17. 

loeblich,  a.  r.,  Jr.  and  loeblich,  a.  r.  hi.  1968.  Index  to  the  genera,  subgenera,  and  sections  of  the 
Pyrrhophyta,  II.  J.  Paleont.  42,  210-213. 

morgenroth,  p.  1970.  Dinoflagellate  cysts  from  the  Lias  Delta  of  Luhnde  Germany.  Neues.  Jb.  Geol.  Palaont. 
Abh.  136,  345-359,  pis.  9-13. 

schrank,  F.  von  p.  1793.  Mikroskopische  Wahrnehmungen.  Der  Naturforscher,  27,  26-37,  pi.  3. 
stein,  f.  r.  von.  1883.  Der  Organismus  der  Infusionsthiere  nach  eigenen  Forschungen  in  systematischer 
Reihenfolge  bearbeitet.  Abt.  3,  Hf.  2.  Die  Naturgeschichte  der  arthrodelen  Flagellaten.  Leipzig.  30  pp.,  25  pis. 
stosch,  h.  a.  von.  1969.  Dinoflagellaten  aus  der  Nordsee  II.  Helgolandinium  subglobosum.  gen.  et  spec.  nov. 
Helgolander  wiss.  Meeresunters,  19,  569-577. 

stover,  L.  E.  and  evitt,  w.  R.  1978.  Analyses  of  pre-Pleistocene  organic-walled  dinoflagellates.  Stanf.  Univ. 
Pubis  (Geol.  Sci.),  15,  1-300. 


688 


PALAEONTOLOGY,  VOLUME  23 


wall,  d.  and  dale,  B.  1970.  Living  hystrichosphaerid  dinoflagellate  spores  from  Bermuda  and  Puerto  Rico. 
Micropaleontology , 16,  47-58,  pi.  1. 

— and  evitt,  w.  r.  1975.  A comparison  of  the  modern  genus  Ceratium  Schrank,  1793,  with  certain  Cretaceous 
marine  dinoflagellates.  Ibid.  21,  14-44,  pis.  1-3. 
wiggins,  V.  d.  1973.  Upper  Triassic  dinoflagellates  from  arctic  Alaska.  Ibid.  19,  1-17,  pis.  1-5. 
williams,  G.  L.  and  downie,  c.  1966.  Further  dinoflagellate  cysts  from  the  London  Clay.  Pp.  215-235.  In 
davey,  r.  j.,  downie,  c.,  sarjeant,  w.  A.  s.,  and  williams,  G.  l.  Studies  on  Mesozoic  and  Cainozoic  dino- 
flagellate cysts.  Bull.  Br.  Mus.  nat.  Hist.  (Geol.),  Suppl.  3,  1-248,  pis.  1-25. 


Typescript  received  12  July  1979 
Revised  typescript  received  25  October  1979 


GEOFFREY  L.  EATON 
BP  Petroleum  Development  Limited 
Farburn  Industrial  Estate 
Dyce,  Aberdeen  AB2  OPB 


MODE  OF  LIFE  OF  A GIANT  CAPULID 
GASTROPOD  FROM  THE  UPPER  CRETACEOUS 
OF  SAGH ALIEN  AND  JAPAN 

by  itaru  hayami  and  YASUMITSU  kanie 


Abstract.  The  life  habits  of  a huge  Campanian  patelliform  gastropod,  hitherto  called  ‘ Helcion  giganteus',  from 
Saghalien  and  Japan  are  discussed  on  the  basis  of  several  specimens  adhering  to  enormous  shells  of  Inoceramus 
( Sphenoceramus ) schmidti.  This  gastropod  is  here  transferred  to  the  Capulidae  of  Mesogastropoda,  and  a new 
generic  name,  Gigantocapulus,  is  proposed  for  it.  Its  ecological  relation  with  7.  (S.)  schmidti  is  regarded  as 
parasitic  by  analogy  to  some  living  species  of  Capulus  that  attach  to  the  valves  of  pectinids.  This  interpretation  is 
supported  by  stratigraphic  and  geographic  distribution  patterns  and  by  its  functional  morphology. 

‘ Helcion  giganteus',  originally  described  by  Schmidt  (1873)  from  the  Upper  Cretaceous  at  Cape  Dui 
near  Alexandrovsk,  north  Saghalien,  is  probably  the  largest  patelliform  gastropod  known.  Its  shell 
sometimes  exceeds  400  mm  in  maximum  length,  and  shows  a wide  range  of  morphological  variation. 
This  species,  though  restricted  to  the  lower  to  middle  Campanian  (Zone  of  Inoceramus  schmidti), 
occurs  at  various  localities  in  Saghalien,  Japan  (mainly  Hokkaido),  Koryak  Highland  of  eastern 
Siberia  (Dundo  and  Efremova  1974),  Southern  Alaska  (Jones,  pers.  comm.),  and  British  Columbia 
(Whiteaves  1903).  The  association  of  this  species  with  I.  ( Sphenoceramus ) schmidti  Michael,  1899,  is 
important.  Almost  all  the  specimens  of  ‘7.  digitatus'  described  by  Schmidt  (1873)  together  with 
‘77.  giganteus ' from  Cape  Dui  seem  to  be  referable  to  7.  (S'.)  schmidti,  as  revised  in  Michael  (1899)  and 
Nagao  and  Matumoto  (1940).  Their  coexistence  in  the  same  fossil  bed  (commonly  fine-grained 
sandstone)  was  also  recorded  at  many  other  localities:  Naibuchi  (=Naibuti)  (Matumoto  1942, 
p.  167)  in  south  Saghalien,  Abeshinai  (Matumoto  1942,  p.  205),  Hetonai  (Matumoto  1942,  p.  251), 
Urakawa  (Matumoto  1942,  p.  268;  Kanie  1966,  p.  322;  1977,  p.  54)  and  some  other  places  in 
Hokkaido,  and  Dogo-Himezuka,  Matsuyama  City  (Kashima  1972;  Matsumoto  1973)  in  Shikoku. 

Summarizing  the  classification  and  evolutionary  history  of  Cretaceous  patelliform  gastropods  in 
the  northern  Pacific  region,  Kanie  (1975)  concluded  that  ‘77.  giganteus ' belongs  to  the  Meso- 
gastropoda and  that  they  possibly  attached  to  some  other  shelled  organism.  Since  ‘77.  giganteus'  is 
seldom  accompanied  by  molluscs  assumed  to  have  lived  on  near-shore  rocky  substrates,  it  was 
assumed  that  it  may  have  been  attached  to  large  bivalves  such  as  7.  ( S .)  schmidti,  but  at  that  time  there 
was  no  direct  evidence.  Subsequently  Hayami  found  a specimen  of ‘77.  giganteus',  in  growth  position 
attached  to  the  shell  surface  of  Inoceramus,  in  the  collection  of  the  University  Museum,  University  of 
Tokyo.  We  have  now  examined  the  relation  between  the  two  molluscs  on  the  basis  of  many  specimens 
stored  at  various  institutions.  In  the  present  article  we  describe  some  of  these  specimens,  discuss  the 
interpreted  life  habit  of  this  gastropod,  and  compare  it  with  some  living  species  of  similar  habit.  The 
taxonomic  position  of ‘77.  giganteus'  is  also  reconsidered. 


SYSTEMATIC  PALAEONTOLOGY 


Order  caenogastropoda  Cox,  1959 
Suborder  mesogastropoda  Thiele,  1925 
Superfamily  calyptraeacea  Lamarck,  1809 
Family  capulidae  Fleming,  1822 
Genus  Gigantocapulus  Hayami  and  Kanie,  gen.  nov. 


IPalaeontology,  Vol.  23,  Part  3, 1980,  pp.  689-698,  pi.  87.1 


690 


PALAEONTOLOGY,  VOLUME  23 


Type  species.  Helcion  giganteus  Schmidt,  1873,  northern  Pacific  region,  Campanian. 

Diagnosis.  Shell  very  large,  cap-shaped  or  conical,  bilaterally  symmetrical  but  more  or  less  irregular  in  outline; 
apex  located  anteriorly  from  the  centre,  sometimes  marginal;  surface  commonly  ornamented  with  irregularly 
disposed  radial  costae  in  addition  to  concentric  rings  on  the  apical  region;  anterior  elevated  sector  and  internal 
septum  absent;  outermost  layer  prismatic,  while  other  and  inner  layers  are  crossed-lamellar;  some  species  living 
upon  the  shells  of  Inoceramus. 

Remarks.  The  taxonomic  position  of  ‘ H . giganteus ’ and  its  allied  species  from  the  Cretaceous  of 
northern  Pacific  has  been  debatable;  Capulus,  Patella,  Scurria,  Acmaea,  and  Brunonia  also  have  been 
used  as  their  generic  names.  Living  patelliform  gastropods  occur  in  various  unrelated  taxonomic 
groups,  e.g.  the  Patellacea  of  Archaeogastropoda,  the  Neritacea  and  Calyptraeacea  of  Meso- 
gastropoda  and  the  Siphonariacea  of  Pulmonata.  Because  their  shell  forms  sometimes  show 
remarkable  convergence,  such  essential  characters  as  muscle  impression,  presence  or  absence  of 
internal  septum  and  shell  structure  as  well  as  inferable  life  habit  may  be  important  for  determination 
of  the  taxonomic  position  of  fossil  species. 

Kanie  (1975)  assigned  these  Cretaceous  species  in  question  to  the  genus  Anisomyon  Meek  and 
Hayden,  1860,  which  had  been  included  in  the  Basommatophora  (an  order  of  Pulmonata),  and 
proposed  a new  family  Anisomyonidae  in  the  Mesogastropoda.  This  treatment  was  primarily  based 
on  the  resemblance  of  muscle  impressions  and  shell  form  of  some  species  to  the  Capulidae  and  the 
difference  of  shell  structure  from  the  Siphonariidae.  As  noted  elsewhere  (Hayami  and  Kase  1977, 
p.  55),  however,  one  of  us  (I.  H.)  doubted  if  the  type  species  of  Anisomyon  [H.  patelliformis  Meek  and 
Hayden,  1856]  should  be  transferred  from  the  Basommatophora  to  the  Mesogastropoda,  and 
presumed  that  ‘ H . giganteus ’ may  represent  an  unnamed  genus  of  the  Capulidae.  This  is  proposed 
here,  which  modifies  the  previous  classification  (Kanie  1975)  of  Cretaceous  patelliform  gastropods 
from  the  northern  Pacific  region. 

Kanie  (1975)  distinguished  two  ‘morphotypes’  in  lH.  giganteus':  type  A is  characterized  by  the 
relatively  small  size,  small  apical  angle,  and  irregularly  noded  ornament,  while  type  B has  relatively 
large  size,  large  apical  angle,  and  almost  persistent  and  not  noded  radial  ribs.  Of  the  originally  figured 
specimens  of  H.  giganteus,  most  individuals  including  the  lectotype  (Schmidt  1873,  pi.  2,  fig.  17, 
designated  by  Kanie  (1975)  as  ‘holotype’)  belong  to  type  B,  and  only  two  small  specimens  (Schmidt 
1873,  pi.  3,  figs.  8, 9)  may  belong  to  type  A.  Numerous  individuals  of  type  A are  preserved  in  various 
Japanese  institutions,  but  none  of  them  actually  shows  any  intimate  relation  to  the  shell  of 
Inoceramus.  All  the  observed  specimens  attached  to  the  surface  of  I.  (S.)  schmidti  belong  to  type  B. 
Moreover,  significant  morphological  differences  are  newly  recognized  between  the  two  ‘morpho- 
types’.  First,  a trace  of  an  internal  septum  is  often  seen  in  type  A (see  Kanie  1975,  p.  9,  fig.  2),  but  has 
never  been  observed  in  type  B.  Secondly,  the  apex  is  always  located  subcentrally  or  even  posteriorly  in 
type  A,  while  it  is  commonly  located  very  anteriorly  or  even  near  the  anterior  margin  in  type  B. 
Thirdly,  a tongue-like  projection,  as  described  later,  occurs  only  in  type  B.  Host-determined  non- 
genetic  variation  is  actually  known  in  a living  capulid  species  (Thorson  1965),  and  dwarf  males  are 
also  seen  in  such  semi-parasitic  gastropods.  Yet,  such  great  differences  of  essential  characters  are 
hardly  explicable  by  individual  variation.  At  present,  we  consider  that  the  two  ‘morphotypes’  belong 
to  different  species,  and  that  the  use  of  the  specific  name  Gigantocapulus  giganteus  should  be  restricted 
to  the  type  B of  Kanie  (1975).  The  specimens  of  type  A seem  to  be  close  to  lA.  transformis'  Dundo  and 
Efremova  (1974)  from  the  Koryak  Highland.  The  presence  of  an  internal  septum  may  suggest  that 
they  belong  to  the  Calyptraeidae. 


EXPLANATION  OF  PLATE  87 

Figs.  1,  2.  Gigantocapulus  giganteus  (Schmidt,  1873).  UMUT  MM5535  attached  to  the  surface  of  Inoceramus 
(, Sphenoceramus ) schmidti  Michael,  1 899.  Loc.  N469,  north-west  of  Miho  (gorge  of  Ryugase),  Naibuchi  area, 
south  Saghalien.  Collected  by  T.  Matsumoto.  1,  upper  view,  x 0-42;  2,  left  lateral  view,  x 0-42.  (See  also 
text-fig.  1.) 


PLATE  87 


hayami  and  kanie,  Cretaceous  patelliform  gastropod 


692 


PALAEONTOLOGY,  VOLUME  23 


In  the  Western  Interior  of  the  United  States  some  specimens  of  Anisomyon  have  also  been  found 
adhering  to  the  shells  of  Inoceramus  (Sohl  1967a).  The  association  may  be  comparable  with  the 
present  case.  According  to  Sohl’s  (19676)  redescription  of  A.  patelliformis  (Meek  and  Hayden,  1856), 
however,  one  of  the  paratypes  reveals  clearly  asymmetric  muscle  impression,  which  resembles  that  of 
Siphonaria,  although  the  posterior  carination  of  Siphonaria- type  is  undeveloped  in  that  species.  No 
specimen  of  G.  giganteus  shows  clear  muscle  impression,  but  Capulus- like  horseshoe-shaped  muscle 
scars  are  recognized  in  C.  cassidarius  Yokoyama,  1890,  which  is  considered  to  be  ancestral  to 
G.  giganteus  (Kanie  1975,  p.  9,  fig.  2).  The  genus  Anisomyon  is  represented  by  much  smaller  species 
without  radial  costae,  and  we  are  now  inclined  to  consider  that  it  is  not  directly  related  to 
Gigantocapulus. 

The  genus  Brunonia  Muller,  1898,  may  be  another  Late  Cretaceous  patelliform  gastropod 
comparable  with  our  new  genus  from  morphological  and  paleoecological  standpoints.  The  genus  was 
generally  referred  to  the  Siphonariidae,  but  in  the  Treatise  (Knight  et  al.  1960)  it  was  doubtfully 
included  in  the  suborder  Patellina.  The  concentrically  ornamented  shell  of  its  type  species  [B.  grandis 
Muller,  1898,  from  the  Santonian  of  Germany]  resembles  the  apical  part  of  G.  giganteus. 
Unfortunately,  Muller’s  original  specimen  of  B.  grandis  is  said  to  have  been  lost,  and  further 
comparative  study  is  now  difficult.  At  present  we  think  that  Gigantocapulus  is  at  least  generically 
separable  from  Brunonia  by  the  developed  radial  costae  on  the  surface.  Judging  from  the  original 
figures  of  B.  grandis , the  apex  is  more  constantly  located  near  the  centre  of  shell,  and  no  projection  is 
developed  on  its  anterior  periphery. 

DESCRIPTION  OF  SELECTED  SPECIMENS 

Among  a large  number  of  specimens  of  G.  giganteus  in  the  collection  of  the  University  Museum,  University  of 
Tokyo  (UMUT),  and  the  Institute  of  Geology  and  Palaeontology,  Tohoku  University,  Sendai  (IGPS),  several 
show  an  intimate  association  with  the  shells  of  I.  ( S .)  schmidti.  UMUT  MM5535  (PI.  87;  text-fig.  1)  has  well- 
preserved  shells  of  the  two  species,  and  shows  the  position  and  orientation  of  attachment.  It  was  found  in 
T.  Matsumoto’s  collection  from  a greenish  fine-grained  sandstone  of  the  Ray  1 Member  of  the  Ryugase  Group 
at  loc.  N469  (gorge  of  Ryugase),  about  4-5  km  north-west  of  Miho,  Naibuchi  area,  south  Saghalien  (see  the 
locality  map  in  Matsumoto  1942).  The  following  description  is  entirely  based  on  this  specimen. 

The  shell  of  G.  giganteus,  though  a considerable  part  of  the  marginal  area  is  broken  off,  exceeds  290  mm  in 
maximum  length  and  250  mm  in  breadth,  showing  a suboval,  nearly  bilaterally  symmetrical,  cap-shaped  outline 
with  a somewhat  irregularly  undulating  marginal  area.  The  apex  is  located  at  about  one-fifth  of  maximum  length 


LV  MRV 


text-fig.  1 . Sketch  of  Specimen  I (UMUT  MM5535)  of  Gigantocapulus  giganteus  (Schmidt)  from  the  left  side. 
LV:  Fracture  of  the  left  valve  (prismatic  outer  layer)  of  Inoceramus  ( Sphenoceramus ) schmidti  Michael;  MRV: 
Internal  surface  of  the  right  valve  of  the  same  individual,  on  which  characteristic  divergent  ribs  are  impressed. 


HAYAMI  AND  KANIE:  GIANT  GASTROPOD 


693 


from  the  anterior  end,  but  the  growth-lines  indicate  that  it  was  situated  near  the  centre  of  shell  in  the  early  growth 
stage.  The  pre-apical  area  steeply  descends  towards  the  anterior  margin,  while  the  post-apical  area  is  widely 
expanded  and  broadly  convex.  The  maximum  inflation  of  shell  lies  far  behind  the  apex.  The  surface  is 
ornamented  with  several  subconcentric  ribs  and  about  fifty  radial  costae.  Subconcentric  ribs  are  distinct  and 
widely  spaced,  and  their  distribution  is  confined  to  the  apical  area  (within  50  mm  from  the  apex).  On  the 
contrary,  radial  costae  are  at  first  indistinct  but  become  prominent  after  the  effacement  of  subconcentric  ribs. 
They  are  commonly  irregularly  dichotomous  but  sometimes  convergent.  The  thickness  of  test  does  not  exceed 
10  mm.  The  outermost  layer  is  thin  and  prismatic,  and  the  outer  and  inner  layers  are  crossed-lamellar  and  solid. 

The  associated  shell  of  Inoceramus  is  evidently  a part  of  an  articulated  individual.  The  shell  of  G.  giganteus 
adheres  closely  to  the  surface  of  its  left  valve.  The  shell  margin  of  G.  giganteus , though  its  right  side  is  incomplete, 
fits  perfectly  the  undulating  surface  of  Inoceramus  without  any  perceptible  gap  (text-fig.  1).  The  opposite  valve  of 
Inoceramus  has  been  almost  entirely  exfoliated  and  lost,  but  the  divergent  ribs  impressed  on  the  internal  surface 
are  unmistakably  characteristic  of  I.  ( S .)  schmidti.  The  radial  ribs  of  G.  giganteus  are  evidently  denser  than  the 
divergent  ribs  of  7.  ( S .)  schmidti;  the  former  does  not  necessarily  correspond  with  the  latter.  Judging  from  the 
orientation  of  the  divergent  ribs  as  well  as  the  nearly  closed  valves  of  I.  ( S .)  schmidti  below  the  anterior  margin  of 
G.  giganteus,  this  gastropod  appears  to  have  sat  on  the  antero-ventral  area  of  the  living  shell  of  7.  (S.)  schmidti 
with  the  apex  located  on  the  antero-ventral  side.  The  axis  of  symmetry  of  G.  giganteus  forms  an  angle  of  about 
30°  with  the  line  of  maximum  length  of  7.  ( S .)  schmidti.  The  prismatic  layer  of  Inoceramus,  which  represents  the 
outer  layer  of  the  shell,  is  about  7-0  mm  and  3-0  mm  thick  below  the  anterior  and  posterior  margins  of 
G.  giganteus,  respectively.  The  original  size  of  this  inoceramid  shell  would  exceed  500  mm,  provided  that  the 
thickness  of  this  layer  increases  isometrically  to  the  attained  shell  length.  If  the  allometric  growth  indices 
calculated  by  Tanabe  (1973,  p.  177)  on  some  specimens  of  7.  ( S .)  schmidti  from  Hokkaido  are  applied,  the 
restored  shell  of  this  individual  may  exceed  700  mm  in  maximum  length. 

UMUT  MM57 1 1 ( = Cr.  1217)  (text-fig.  2)  is  interesting  because  its  right-anterior  margin  is  nearly  complete.  It 
belongs  to  an  old  collection  from  the  Cape  Khoi  Beds  at  Cape  Jonquiere  near  Alexandrovsk,  north  Saghalien. 


text-fig.  2.  Gigantocapulus  giganteus  (Schmidt).  Specimen  II  (UMUT  MM5711)  attached  to  a 
crushed  shell  of  Inoceramus  ( Sphenoceramus ) schmidti  Michael.  Loc.  Cape  Jonquiere  near 
Alexandrovsk,  north  Saghalien.  a,  upper  view,  x 0-36;  b,  bird’s-eye  anterior  view  of  the  anterior 
part  of  the  specimen  showing  a tongue-like  projection  and  nearly  complete  right-anterior  margin 
of  shell,  x 0-55;  c,  anterior  view  of  the  same  specimen,  xO-55. 


694 


PALAEONTOLOGY,  VOLUME  23 


This  individual  is  also  closely  associated  with  an  enormous  articulated  shell  of  I.  (S.)  schmidti,  which,  however,  is 
so  strongly  crushed  that  the  original  state  of  attachment  is  difficult  to  restore. 

This  specimen  is  about  250  mm  long  and  190  mm  wide,  and  the  shell  of  post-apical  part  has  been  considerably 
exfoliated  and  lost.  The  matrix  was  successfully  removed  from  the  pre-apical  part  of  shell,  and  both  the  external 
and  internal  characters  are  well  exhibited.  The  marginal  area  of  the  pre-apical  part  is  remarkably  depressed  and 
gently  folded  like  a brim  (text-fig.  2c).  Furthermore,  there  is  a curious  tongue-like  projection  at  the  anterior 
extremity,  which  is  unusually  thickened  with  a rounded  edge.  The  internal  surface  is  nearly  smooth,  and  neither  a 
septum  nor  a muscle  scar  is  observed  below  the  apical  area.  Radial  costae  are  not  impressed  on  the  internal 
surface  even  near  the  margin. 

The  following  specimens  of  G.  giganteus  are  also  intimately  associated  with  some  crushed  shells  of  /.  (S'.) 
schmidti:  UMUT  MM5710  (=Cr.l418)  and  UMUT  MM5709  (=Cr.998):  old  collection  from  the  Zone  of 
I.  schmidti  in  Naibuchi  area,  south  Saghalien  (exact  locality  unknown).  UMUT  MM5713  (=Cr.l228):  old 
collection  from  the  same  locality  as  UMUT  MM5711.  UMUT  MM5712  (=  Cr.1218):  old  collection  from  the 
Zone  of  I.  schmidti  in  Alexandrovsk  area,  north  Saghalien  (exact  locality  unknown).  These  specimens  show  a 
wide  range  of  morphological  variation.  One  of  the  illustrated  paralectotypes  of  Helcion  giganteus  from  the  type 
locality  (Schmidt  1873,  pi.  3,  fig.  2)  may  be  another  example  of  an  attached  specimen,  because  a fragmentary 
prismatic  shell  was  indicated  below  it. 

VARIABILITY  OF  SOME  MORPHOLOGICAL  CHARACTERS 

When  H.  giganteus  was  originally  described  by  Schmidt  (1873),  four  varieties  were  distinguished  by 
the  different  position  of  the  apex.  All  the  original  specimens  are  included  either  in  var.  a depressa,  var. 
j8  nasuta,  var.  y retracta  or  var.  8 centralis.  Kanie  (1975,  p.  23)  designated  a specimen  of  var.  depressa 
as  the  ‘holotype’  of  H.  giganteus,  but  (Hayami  and  Kase  1977,  p.  56)  this  procedure  can  be  regarded 
as  constituting  valid  lectotype  designation.  Dundo  and  Efremova  (1974)  regarded  some  of  these 
varieties  as  distinct  species,  and  referred  centralis  and  nasutus  to  Patella  and  Helcion,  respectively. 
However  (Kanie  1975;  Hayami  and  Kase  1977),  none  of  these  varieties  (except  for  two  small 
specimens  of  var.  depressa  (Schmidt  1873,  pi.  3,  figs.  8, 9)  seems  to  constitute  a distinct  taxon,  because 
the  difference  of  apical  position  as  well  as  other  characteristics  appears  to  be  gradational  within  a 
single  fossil  population.  The  growth-lines  of  the  present  specimens  show  that  the  variability  of  apical 
position  is  partly  due  to  ontogenetic  transformation:  the  apex  evidently  shifts  from  the  central  part  to 
the  anterior  portion  of  shell  with  growth.  There  is  also  a change  of  the  direction  of  apex  in  the  young 
stage.  As  shown  in  UMUT  MM  5709  and  some  other  small  specimens,  the  very  apex,  if  preserved, 
seems  to  point  in  the  direction  opposite  to  the  expansion  of  shell.  Although  the  apex  is  generally 
located  posteriorly  in  many  living  species  of  Capulus  and  related  genera,  this  ontogenetic  change  is 
one  of  the  main  reasons  why  we  suspect  here,  unlike  a previous  interpretation  (Kanie  1975),  that  the 
shorter  end  is  actually  anterior. 

The  shell  form  and  surface  ornamentation  are  also  quite  variable.  Among  the  forty  specimens  we 
have  observed  at  various  institutions  in  Japan,  the  angle  of  ultimate  apex  in  lateral  view  varies  from 
120°  to  145°.  The  number  of  radial  costae  ranges  from  thirty-five  to  sixty-five.  The  thickened  tongue- 
like projection  at  the  anterior  end  of  shell  is  also  observable  in  some  other  specimens,  e.g.  one  of  the 
paralectotypes  (Schmidt  1873,  pi.  3,  fig.  10)  and  IGPS  no.  50910  (Kanie  1975,  pi.  15,  fig.  1 a,  b\  Kanie 
1977,  pi.  2,  fig.  4).  It  may  be  a widespread  character  in  this  species,  but,  as  shown  by  the  growth-lines 
on  UMUT  MM5535,  5711,  its  development  is  seen  only  in  the  later  ontogenetic  stage. 

The  range  of  morphological  variation  of  G.  giganteus  is  thus  unusually  wide.  Such  a great 
variability  is  unknown  in  any  living  species  of  the  Patellacea,  but  comparable  with  that  of  some 
species  of  the  Capulidae.  The  variable  shell  form  and  ornamentation  of  this  species  were  probably 
influenced  by  the  nature  of  the  surface  of  the  host. 

INTERPRETATION  OF  MODE  OF  LIFE 

From  our  observation  on  in  situ  specimens,  it  is  likely  that  at  least  some  individuals  of  G.  giganteus 
grew  on  living  shells  of  Inoceramus  ( Sphenoceramus ) schmidti.  This  is  supported  by  the  fact  that  the 
associated  inoceramid  shells  are,  even  if  crushed,  commonly  articulated.  Moreover,  the  stratigraphic 


HAYAMI  AND  KANIE:  GIANT  GASTROPOD 


695 


and  geographic  distribution  of  the  two  species  is  identical,  which  suggests  not  only  their  intimate 
ecological  relation  but  also  that  the  evolutionary  history  of  the  former  depended  on  the  latter. 

A large  number  of  malacologists  and  marine  ecologists  have  paid  attention  to  the  parasitic  or  semi- 
parasitic  life  of  Capulus  species  and  their  hosts  (Orton  1912,  1949;  Yonge  1938;  Otuka  1939; 
Teramachi  1942;  Kuroda  1951;  Sharman  1956;  Burch  and  Burch  1961;  Orr  1962;  Thorson  1965; 
Kosuge  and  Hayashi  1967;  Habe  1967).  The  hosts  are  commonly  epifaunal  bivalves,  especially  large 
species  of  the  Pectinacea,  though  in  a few  cases  the  epibionts  rest  also  on  the  surface  of  certain 
gastropods,  brachiopods,  and  annelids.  Sometimes  an  almost  exclusive  relation  exists  between  the 
epibiont  and  host  species  (e.g.  C.tosaensis  on  Propeamussium  sibogae  in  Japan),  but  in  other  cases  an 
epibiont  species  can  grow  on  various  hosts  (e.g.  C.  dilatatus  on  Amusium  japonicum,  Pecten  albicans, 
Decatopecten  striatus,  Chlamys  nobilis,  etc.  in  Japan;  C.  ungaricus  on  P.  maximus,  Aequipecten 
opercularis.  Modiolus  modiolus,  Monia  patelliformis,  Turritella  communis,  etc.  in  Great  Britain  and 
North  Sea).  According  to  Yonge  (1938)  and  others,  C.  ungaricus  is  a ciliary  feeder.  It  intercepts  the 
food,  which  has  been  collected  on  the  gills  of  a bivalve,  by  inserting  its  long  proboscis  inside  the 
bivalve  shells.  The  ecological  relation  was  regarded  as  semi-parasitic  by  Sharman  (1956)  and  as 
commensalistic  by  Thorson  (1965).  Although  the  epibiont  does  not  seem  to  cause  the  bivalves  any 
mortal  harm,  this  state  is  most  certainly  disadvantageous  to  the  host.  We  consider  that  this  is  a case  of 
external  parasitism,  but  the  term  ‘semi-parasitic’  may  be  more  appropriate  for  this  species,  because  it 
also  attaches  to  dead  shells  and  rocks. 

The  life  habits  of  such  parasitic  individuals  of  Capulus  can  be  classified  into  two  types.  One  is 
represented  by  C.  dilatatus,  in  which  (Kosuge  and  Hayashi  1967)  the  epibiont  bores  a small  hole 
through  the  pectinid  shell  (commonly  up-facing  valve)  in  order  to  insert  its  proboscis.  The  boring 
position  is  concentrated  on  the  anterior  half  of  the  disc  (corresponding  to  the  position  of  gills)  and 
sometimes  on  the  anterior  wing.  The  orientation  of  attachment  seems  to  be  almost  random.  The  other 
type  is  exemplified  by  C.  ungaricus,  which  rests  preferentially  on  the  anterior  and  ventral  marginal 
part  of  down-facing  valves  of  living  pectinids.  Sharman  (1956)  examined  the  attaching  position  and 
orientation  of  many  individuals  of  this  species  on  the  shells  of  A.  opercularis  from  off  the  coast  of  the 
Isle  of  Man,  noting:  ‘in  its  characteristic  position  the  gastropod  sits  at  the  edge  of  the  valve  with  the 
front  margin  of  the  shell  projecting  a little  over  it  and  the  apex  pointing  inwards.’  C.  ungaricus  never 
makes  a borehole,  but  the  edge  of  the  valve  margin  is  said  to  be  frequently  chipped  so  that  this 
gastropod  can  easily  insert  its  proboscis  into  the  pectinid  valves.  Somewhat  similar  feeding  habits  are 
known  in  C.  tosaensis  from  the  Japanese  deep  waters,  although  this  species  is  said  to  attach 
preferentially  to  the  left  (?  up-facing)  valve  of  Propeamussium  sibogae. 

On  the  shell  surface  of  the  many  specimens  of  I.  ( S .)  schmidti  neither  a borehole  nor  a scar  of 
attachment  has  been  recognized,  and  it  may  be  difficult  to  know  whether  the  valve  margin  was 
actually  chipped  or  not  by  other  organisms.  However,  the  attaching  position  and  orientation  in  the 
specimens  on  text-fig.  1 seem  to  indicate  that  the  life  habit  of  G.  giganteus  was  analogous  to  the  second 
type,  especially  to  the  case  of  C.  ungaricus  as  illustrated  by  Sharman  ( 1 956,  figs.  1-3).  We  interpret  the 
function  of  the  curious  tongue-like  anterior  projection  as  protecting  the  head  of  the  gastropod  which 
presumably  protruded  a little  beyond  the  edge  of  the  valve  margin,  because  otherwise  the  remarkable, 
declined  margin  of  this  projection  could  not  adhere  closely  to  the  surface  of  inoceramid  shell.  Such  a 
hanging  front  margin  of  the  shell  is  also  commonly  seen  in  C.  ungaricus.  Text-fig.  3 shows  a putative 
living  position  of  G.  giganteus  on  the  left  valve  of  I.  ( S .)  schmidti,  although  it  is  still  unknown  whether 
the  valve  is  actually  up-facing  or  down-facing. 


SUMMARY 

The  observation  of  in  situ  specimens  and  the  functional  interpretation  of  the  shell  shows  that 
G.  giganteus  was  a parasitic  gastropod  to  I.  ( S .)  schmidti.  Considering  the  much  smaller  size  of  other 
associated  molluscs,  only  this  inoceramid  seems  to  have  offered  the  solid  ground  of  attachment  for 
such  a large  patelliform  gastropod.  Although  complete  specimens  of  I.  ( S .)  schmidti  can  seldom  be 


696 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  3.  Reconstruction  of  the  living  position  of  Gigantocapulus 
giganteus  (Schmidt)  on  Inoceramus  ( Sphenoceramus ) schmidti  Michael. 

Their  periostracum  is  not  drawn,  because  nothing  is  known  about  its 
development.  This  is  not  a sketch  but  chiefly  based  on  Specimens  I and  II. 

obtained,  we  have  actually  observed  several  extraordinarily  large  specimens  of  this  species  (exceeding 
700  mm  in  maximum  length)  in  the  collections  from  Saghalien  and  Hokkaido.  The  gigantism  of  this 
gastropod  is  evidently  related  to  the  unusually  large  size  of  the  host.  If  such  a parasitic  relation  was 
developed,  it  can  be  readily  imagined  that  an  ecologically  specialized  epibiont  was  compelled  to 
become  extinct  by  the  decline  of  the  host  species.  G.  giganteus  seems  to  have  shared  its  evolutionary 
lot  with  I.  ( S .)  schmidti,  because  their  stratigraphic  and  geographic  distribution  is  identical. 

The  history  of  this  external  parasitism  possibly  goes  back  to  earlier  times.  As  interpreted  previously 
(Kanie  1975),  G.  giganteus  may  have  been  derived  from  C.  cassidarius  Yokoyama  through  some 
intermediate  form.  C.  cassidarius  is  common  in  the  Turonian  to  Santonian  strata  of  the  same  region 
and  is  frequently  accompanied  by  I.  ( S .)  naumanni  Yokoyama,  which  seems  to  be  ancestral  to  I.  (S'.) 


HAYAMI  AND  KANIE:  GIANT  GASTROPOD 


697 


schmidti,  and  some  other  small-sized  species  of  Inoceramus.  Therefore,  it  is  possible  that  the 
parasitism  was  already  established  between  the  ancestors,  although  in  situ  preservation  has  not  been 
found.  The  inferred  mode  of  life  also  explains  shell  orientation,  the  wide  range  of  morphological 
variation,  and  the  curious  tongue-like  anterior  projection  in  G.  giganteus.  Its  taxonomic  reference  to 
the  Capulidae  of  Mesogastropoda  is  also  consistent  with  the  parasitic  mode  of  life.  There  is  still  a 
shortage  of  in  situ  material  showing  life  orientations,  which  will  provide  more  evidence  of  the 
paleoecological  relation  between  this  peculiar  limpet  and  other  organisms. 

Acknowledgements.  We  are  grateful  to  Dr.  Norman  F.  Sohl  (U.S.  Geological  Survey)  for  his  helpful  suggestions, 
reading  the  manuscript,  and  loan  of  Anisomyon;  to  Professor  Emeritus  Tatsuro  Matsumoto  (Kyushu 
University)  for  the  Saghalien  specimens.  Professor  Tamio  Kotaka  and  Dr.  Kenshiro  Ogasawara  (Tohoku 
University)  made  available  material  at  their  institute.  We  also  thank  Professor  Tetsuro  Hanai,  Professor 
Masuoki  Horikoshi,  and  Mr.  Paul  Frydl  (University  of  Tokyo),  and  Dr.  Masayuki  Tashiro  (Kochi  University) 
for  drawing  text-figure  3. 


REFERENCES 

burch,  J.  Q.  and  burch,  r.  l.  1961.  A new  Capulus  from  Gulf  of  California.  Nautilus , 75,  19-20. 
dundo,  o.  P.  and  efremova,  v.  i.  1974.  [Field  Atlas  of  the  Senonian  Index  Fauna  in  the  Northeastern  Part  of 
Koryak  Highland .]  Sci.  Res.  Inst.  Arct.  Geol.  (NIIGA),  Leningrad.  28  pp.,  12  pis.  [In  Russian.] 
habe,  T.  1967.  A new  capulid  snail,  Capulus  spondylicola,  from  Japan.  Venus,  26,  (2),  37-38,  pi.  4. 
hayami,  i.  and  kase,  t.  1977.  A systematic  survey  of  the  Paleozoic  and  Mesozoic  Gastropoda  and  Paleozoic 
Bivalvia  from  Japan.  Univ.  Mus.  Univ.  Tokyo,  Bull.  13,  1-155. 
kanie,  y.  1966.  The  Cretaceous  deposits  in  the  Urakawa  district,  Hokkaido.  J.  Geol.  Soc.  Japan,  72  (7),  315-328. 
[In  Japanese  with  English  abstract.] 

— 1975.  Some  Cretaceous  patelliform  gastropods  from  the  northern  Pacific  region.  Sci.  Rept.  Yokosuka  City 
Mus.  21,  1-44,  pis.  1-20. 

— 1977.  Succession  of  the  Cretaceous  patelliform  gastropods  in  the  northern  Pacific  region.  In  matsumoto,  t. 
[org.]:  Mid-Cretaceous  Events— Hokkaido  Symposium,  1976.  Palaeontol.  Soc.  Japan,  Spec.  Papers,  21, 
53-62,  pi.  2. 

kashima,  n.  1972.  [Fossils  from  the  Dogo-Himezuka,  Matsuyama  City,  No.  2.]  Ehime  no  Shizen,  14  (8),  12-15. 
[In  Japanese.] 

knight,  J.  B.  et  al.  1960.  Treatise  on  Invertebrate  Paleontology.  Part  I.  Mollusca  1.  Geol.  Soc.  America  and  Univ. 
Kansas  Press.  351  pp. 

kosuge,  s.  and  hayashi,  s.  1967.  Notes  on  the  feeding  habits  of  Capulus  dilatatus  A.  Adams,  1 860  (Gastropoda). 

Sci.  Rept.  Yokosuka  City  Mus.  13,  45-54,  pis.  1,  2.  [In  Japanese  with  English  abstract.] 
kuroda,  T.  1951.  [ Capulus  tosaensis  living  on  Propeamussium  sibogae .]  Yumehamaguri,  6,  19.  [In  Japanese.] 
matsumoto,  T.  1973.  Further  notes  on  the  Dogo-Himezuka  fauna.  J.  Geol.  Soc.  Japan,  79  (7),  496.  [In  Japanese.] 
matumoto  [matsumoto],  t.  1942.  Fundamentals  in  the  Cretaceous  stratigraphy  of  Japan.  Mem.  Fac.  Sci. 
Kyushu  Univ.  [D]  1 (3),  129-280,  pis.  5-20. 

meek,  f.  b.  and  hayden,  f.  v.  1857.  Description  of  a new  species  of  Gastropoda  from  the  Cretaceous  formations 
of  Nebraska  Territory.  Proc.  Philad.  Acad.  Nat.  Sci.,  for  1856,  8,  63-69. 

— 1860.  Systematic  catalogue  with  synonymy  of  Jurassic  and  Cretaceous,  and  Tertiary  fossils  collected 
in  Nebraska.  Ibid.  12,  412-432. 

Michael,  r.  1899.  Ueber  Kreidefossilien  von  der  Insel  Sakhalin.  Jahrb.  k.  preuss.  geol.  Landesanst.  18,  153-164, 
pis.  5,  6. 

Muller,  G.  1898.  Die  molluskenfauna  des  Untersenon  von  Braunschweig  und  Ilsede.  I.  Lamellibranchiaten  und 
Glossphoren.  Abh.  preuss.  geol.  Landes,  n.f.,  25,  1-142,  pis.  1-18. 
nago,  t.  and  matumoto  [matsumoto],  t.  1940.  A monograph  of  the  Cretaceous  Inoceramus  of  Japan.  Part  2. 

J.  Fac.  Sci.  Hokkaido  Imp.  Univ.  [4]  6,  1-64,  pis.  1-22. 
orr,  v.  1962.  The  drilling  habit  of  Capulus  danielli.  Veliger,  5,  63-67. 

orton,  j.  h.  1912.  The  mode  of  feeding  of  Crepidula,  with  an  account  of  the  current-producing  mechanism  in  the 
mantle  cavity,  and  some  remarks  on  the  mode  of  feeding  in  gastropods  and  lamellibranchs.  J.  Mar.  Biol.  Ass. 
U.K.  9,  444-478. 

— 1949.  Notes  on  the  feeding  habit  of  Capulus  ungaricus.  Rept.  Mar.  Biol.  Sta.  Pt.  Erin,  61,  29-30. 


698 


PALAEONTOLOGY,  VOLUME  23 


otuka,  Y.  1939.  Non-sculptured  species  of  the  genus  Capulus.  Venus , 9,  89-98,  pi.  4. 

schmidt,  M.  F.  1873.  Ueber  die  Petrefakten  der  Kreideformation  von  der  Insel  Sachalin.  Mem.  Acad.  Imp.  Sci. 
St.-Petersb.,  ser.  7,  19  (3),  1-37,  pis.  1-8. 

sharman,  M.  1956.  Note  on  Capulus  ungaricus  (L.).  J.  Mar.  Biol.  Ass.  U.K.  35,  445-450. 
sohl,  n.  f.  1967 a.  Upper  Cretaceous  gastropod  assemblages  of  the  Western  Interior  of  the  United  States,  in 
Paleoenvironments  of  the  Cretaceous  Seaway — a symposium,  Colorado  School  of  Mines,  Golden,  1-37. 

— 1967 b.  Upper  Cretaceous  gastropods  from  the  Pierre  Shale  at  Red  Bird,  Wyoming.  Prof.  Pap.  U.S.  geol. 
Surv.  393-B,  1-46,  pis.  1-11. 

tanabe,  K.  1973.  Evolution  and  mode  of  life  of  Inoceramus  ( Sphenoceramus ) naumanni  Yokoyama  emend.,  an 
Upper  Cretaceous  bivalve.  Trans.  Proc.  Palaeontol.  Soc.  Japan,  [n.s.]  92,  163-184,  pis.  27,  28. 
teramachi,  a.  1942.  [Ecology  of  Capulus  tosaensis .]  Venus,  12  (1-2),  106.  [In  Japanese.] 
thorson,  G.  1965.  A neotenous  dwarf-form  of  Capulus  ungaricus  (L.)  (Gastropoda,  Prosobranchia)  com- 
mensalistic  on  Turritella  communis  Risso.  Ophelia,  2,  175-210. 
whiteaves,  j.  f.  1903.  On  some  additional  fossils  from  the  Vancouver  Cretaceous,  with  a revised  list  of  the 
species  therefrom.  Geol.  Surv.  Canada,  Mesozoic  Fossils,  1 (5),  309-416,  pis.  40-51. 
yokoyama,  M.  1890.  Versteinerungen  aus  der  japanischen  Kreide.  Palaeontographica,  36,  159-202,  pis.  18-25. 
yonge,  c.  M.  1938.  Evolution  of  ciliary  feeding  in  the  Prosobranchia,  with  an  account  of  feeding  in  Capulus 
ungaricus.  J.  Mar.  Biol.  Ass.  U.K.  22,  453-468. 


ITARU  HAYAMI 

University  Museum 
University  of  Tokyo 
Hongo  7-3-1,  Bunkyo-ku 
Tokyo  113,  Japan 


YASUMITSU  KANIE 


Typescript  received  19  July  1979 

Revised  typescript  received  20  November  1979 


Yokosuka  City  Museum 
Fukadadai  95 
Yokosuka  238,  Japan 


TWO  NEW  JURASSIC  BRYOZOA  FROM 
SOUTHERN  ENGLAND 

by  PAUL  D.  TAYLOR 


Abstract.  Reptomultisparsa  tumida  sp.  nov.  and  Reptoclausa  porcata  sp.  nov.  are  described  respectively  from 
the  Bathonian  Bradford  Clay  of  Bradford-on- Avon  and  the  Aalenian/Bajocian  Inferior  Oolite  of  the  Cotswold 
Hills.  The  genus  Reptoclausa  was  previously  known  only  from  the  Cretaceous.  Reptoclausa  colonies  have 
autozooecia  located  on  longitudinal  ridges  which  are  separated  by  furrows  composed  of  kenozooecia.  This 
unusual  arrangement  of  zooecia  can  be  explained  by  the  action  of  physiological  growth  gradients  during  colony 
life. 

During  a revision  of  some  Jurassic  Bryozoa  from  England  and  Normandy  (Taylor  1977)  two  new 
species  were  recognized  and  are  described  for  the  first  time  in  this  paper.  Both  species  are  encrusting 
forms  belonging  to  the  Order  Cyclostomata,  the  most  abundant  group  of  bryozoans  during  a period 
of  relatively  low  species  diversity.  A tentative  estimate  of  known  Jurassic  bryozoan  diversity  based 
on  available  data  suggests  the  existence  of  about  ninety  species  belonging  to  approximately  thirty 
genera,  although  these  figures  are  undoubtedly  an  underestimate  of  total  worldwide  diversity  for 
three  main  reasons.  First,  almost  all  known  Jurassic  bryozoans  have  been  described  from  either 
England,  France,  or  Germany;  they  are  extremely  poorly  known  from  other  parts  of  the  world  (see 
Taylor  1977,  pp.  336-338).  Secondly,  high  levels  of  phenotypic  variation  within  species  of  simple 
morphology  has  hindered  taxonomic  discrimination  and  quite  probably  has  led  to  undersplitting. 
Genetic  studies  (e.g.  Thorpe  et  al.  1978)  of  living  bryozoans  are  beginning  to  reveal  the  presence  of 
‘cryptic’  species  that  are  difficult  to  distinguish  morphologically.  Thirdly,  non-calcified  ctenostome 
bryozoans  with  a low  fossilization  potential  were  perhaps  much  more  common  during  the  Jurassic 
than  is  immediately  obvious  (Voigt  1977;  Pohowsky  1978;  Taylor  1978).  Despite  their  low  apparent 
diversity  Jurassic  bryozoans  are  ubiquitous  in  marine  sediments,  which  accumulated  in  aerobic 
environments  containing  firm  substrates  (e.g.  brachiopod  shells)  suitable  for  colony  attachment. 
Some  species  developed  erect  growth  from  an  attached  base  but  the  commonest  species  were  totally 
encrusting,  as  are  the  two  new  species  described  here. 

Type  and  figured  specimens  are  housed  in  the  British  Museum  (Natural  History)  (BMNH). 


FAMILIAL  CLASSIFICATION  OF  JURASSIC  TUBULOPORINID  BRYOZOA 

The  order  Cyclostomata  was  represented  in  the  Jurassic  by  two  suborders,  the  Tubuloporina  and  the 
Cerioporina.  Six  major  Jurassic  families  of  the  Tubuloporina  may  be  distinguished:  Stomatoporidae, 
Oncousoeciidae,  Macroeciidae  ( = Multisparidae),  Plagioeciidae,  Theonoidae,  and  Frondiporidae. 
The  stomatoporids  are  typically  encrusting  uniserial  or  narrow  multiserial  (‘ribbon-shaped’)  genera, 
apparently  lacking  the  larval  brooding  polymorphs  known  as  gonozooids  which  characterize  the 
other  tubuloporinid  families.  Oncousoecids  have  branching  adnate  colonies  and  gonozooids  with 
minute  ooeciopores.  Macroecid  and  plagioecid  genera  developed  a variety  of  convergent  colony 
forms,  both  erect  (e.g.  ‘ Entalophora' , ‘ PustuloporcC)  and  encrusting  (e.g.  ‘ Berenicea'),  but  the  two 
families  may  be  distinguished  from  one  another  by  the  structure  of  their  gonozooids.  Macroecid 
gonozooids  are  longitudinally  elongate  and  possess  comparatively  large  ooeciopores  (the  orifice 
through  which  the  larvae  were  released).  Plagioecid  gonozooids  are  broad,  bulbous  and  possess 

[Palaeontology,  Vol.  23,  Part  3,  1980,  pp.  699-706,  pi.  88.| 


700 


PALAEONTOLOGY,  VOLUME  23 


ooeciopores  which  are  considerably  smaller  than  the  apertures  of  autozooids  in  the  same  colony.  In 
both  the  theonoids  and  the  frondiporids,  groups  of  contiguous  autozooidal  apertures  form  fascicles 
elevated  above  the  general  level  of  the  colony  surface.  Frondiporid  zooids  are  typically  longer  than 
those  of  theonoids  and,  whereas  zooidal  budding  occurred  within  the  lengthening  frondiporid 
fascicles,  within-fascicle  zooidal  budding  is  not  known  in  the  theonoids. 


SYSTEMATIC  PALAEONTOLOGY 

Phylum  bryozoa  Ehrenberg,  1831 
Class  stenolaemata  Borg,  1926 
Order  cyclostomata  Busk,  1852 
Suborder  tubuloporina  Milne-Edwards,  1838 
Family  macroeciidae  Canu,  1918 
Genus  reptomultisparsa  d’Orbigny,  1853 

(see  Walter  1969,  p.  75,  for  a revised  generic  diagnosis) 


Reptomultisparsa  tumida  sp.  nov. 

Plate  88,  fig.  1;  text- fig.  1 

Derivation  of  name.  The  trivial  name  tumida  refers  to  the  broad,  swollen  appearance  of  the  gonozooecia. 
Holotype.  BMNH  D 13346  Bathonian,  Bradford  Clay  ( discus  Zone),  Bradford-on-Avon,  Wiltshire. 

Paratypes.  BMNH  D52651a-c,  Bathonian,  Bradford  Clay,  locality  unknown.  Other  specimens  in  the  authors 
collection  are  from  the  Bathonian  White  Limestone  Formation  of  Foss  Cross,  Gloucestershire. 

Diagnosis.  Reptomultisparsa  with  delicate  unilamellar  zoaria;  autozooecia  with  maximum  width 
midway  along  their  frontal  walls  and  possessing  small  apertures;  gonozooecia  broad  and  inflated. 

Description.  Zoaria  (PI.  88,  fig.  1)  are  unilamellar,  fan-shaped,  or  discoidal  (bereniciform).  Zooecia  arise  at 
divisions  of  existing  interzooecial  walls  on  a basal  lamina. 

Autozooecia  have  moderately  long  frontal  walls  characteristically  attaining  maximum  width  midway  along 
their  length.  Interzooecial  walls  form  conspicuous  traces  on  the  relatively  fiat  zoarial  surface.  The  small,  circular 
autozooecial  apertures  are  widely  spaced  and  have  raised  rims  but  lack  distinct  peristomes. 

Kenozooecia  may  occur  around  gonozooecial  borders.  Their  proximal  portions  are  identical  to  those  of 
autozooecia  but  the  kenozooecia  are  truncated  distally  by  gonozooecial  dilation  and  consequently  lack  an 
aperture. 

Gonozooecia  (text-fig.  1)  have  narrow  proximal  portions  and  well-defined  inflated  distal  portions  with  a 
circular  to  oval  shape.  The  sub-terminal  ooeciopores  lack  ooeciostomes  and  are  transversely  elongate  and 
slightly  smaller  than  autozooecial  apertures. 


EXPLANATION  OF  PLATE  88 

Fig.  1.  Reptomultisparsa  tumida  sp.  nov.  BMNH  D13346,  Holotype  colony,  Bathonian,  Bradford  Clay, 
Bradford-on-Avon,  x 7. 

Figs.  2-6.  Reptoclausa  porcata  sp.  nov.  2,  BMNH  B4855,  immature  colony  prior  to  ridge  development, 
Bajocian,  Lower  Ragstone,  Cold  Comfort,  x 7.  3-5,  BMNH  D8724,  Holotype,  Aalenian,  Pea  Grit,  Birdlip. 
3,  intracolony  overgrowth  by  a new  layer  of  zooecia,  x 1 1.  4,  furrow  occupied  by  kenozooecia  (centre  right), 
x 13.  5,  autozooecia  with  rounded  distal  terminations,  x30.  6,  BMNH  D10091,  zoarium  showing 

conspicuous  furrows  and  ridges  with  a ridge  dichotomy,  Aalenian,  Pea  Grit,  Crickley  Hill,  x 8. 

Figs.  1 and  2 are  of  ammonium  chloride  coated  specimens. 


PLATE  88 


taylor,  Jurassic  Bryozoa 


702 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  1.  Reptomultisparsa  tumida  sp.  nov. 
BMNH  D13346  (holotype),  Upper  Bathonian, 
Bradford  Clay,  Bradford-on-Avon,  Wiltshire. 
A group  of  autozooecia  and  a bulbous  gono- 
zooecium.  x 38. 


Dimensions.  See  Table  1. 

Remarks.  The  relatively  broad  and  inflated  gonozooecia  of  Reptomultisparsa  tumida  distinguish  it 
from  other  species  in  the  genus,  and  the  subterminal  position  of  the  ooeciostome  contrasts  with  the 
terminal  ooeciostomes  developed  in  plagioecids  with  similar  bereniciform  colonies. 

Stratigraphical  range.  Upper  Bathonian. 


table  1.  Dimensions  (in  mm)  of  Reptomultisparsa  tumida  zooecia. 
Abbreviations  of  morphological  characters:  law,  longitudinal  auto- 
zooecial  aperture  width;  taw,  transverse  autozooecial  aperture  width;  fwl, 
autozooecial  frontal  wall  length;  fww,  autozooecial  frontal  wall  width 
(maximum);  tgl,  total  gonozooecial  frontal  wall  length;  igl,  length  of 
inflated  portion  of  the  gonozooecial  frontal  wall;  gw,  gonozooecial 
frontal  wall  width  (maximum);  low,  longitudinal  ooeciopore  width;  tow, 
transverse  ooeciopore  width.  Abbreviations  of  statistical  functions:  Nc, 
number  of  colonies  from  which  measurements  were  taken;  Nz,  number  of 
zooecia  measured;  x,  mean  value;  Rc,  range  of  colony  means;  Rz,  total 
range  of  values. 


Nc 

Nz 

X 

Rc 

Rz 

law 

3 

55 

008 

008 

0-06-0-10 

taw 

3 

55 

008 

008 

0-06-0-09 

fwl 

3 

55 

0-67 

0-63-0-69 

0-46-0-88 

fww 

3 

55 

018 

017-019 

0 14-0  21 

tgl 

3 

3 

1-44 

1-20-1-82 

1-20-1-82 

igl 

3 

6 

Ml 

0-92-1-34 

0-86-1-34 

gw 

3 

6 

0-62 

0-53-0-67 

0-51-0-80 

low 

3 

5 

006 

0-06 

0-05-007 

tow 

3 

5 

0-07 

0-06-0-08 

006-0- 10 

TAYLOR:  JURASSIC  BRYOZOA 


703 


Genus  reptoclausa  d’Orbigny,  1853 
Reptoclausa  porcata  sp.  nov. 

Plate  88,  figs.  2-6;  text-fig.  2 

71894  Berenicea  allaudi  (Sauvage);  Gregory,  p.  60. 

1896a  Berenicea  Allaudi  (Sauvage);  Gregory,  p.  44  (partim .). 

1896ft  Berenicea  allaudi  (Sauvage);  Gregory,  p.  77  (partim.),  pi.  3,  fig.  6. 

1969  Idmonea  triquetra  Lamouroux;  Walter,  p.  52  (partim.),  pi.  3,  figs.  11-13  only. 

Derivation  of  name.  The  trivial  name  porcata  refers  to  the  ridged  and  furrowed  form  of  the  zoarium. 

Holotype.  BMNH  D8724,  Aalenian,  Pea  Grit  (murchisonae  Zone),  Birdlip,  Gloucestershire. 

Paratypes.  BMNH  D7526a-b,  Aalenian,  Pea  Grit  (murchisonae  Zone),  near  Stroud,  Gloucestershire.  BMNH 
D31586,  Aalenian,  Lower  Limestone  (murchisonae  Zone),  Crickley  Hill,  Gloucestershire.  BMNH  B2290a-c, 
Inferior  Oolite,  Crickley  Hill,  Gloucestershire.  BMNH  B4855,  Lower  Ragstone  (discites  Zone),  Cold  Comfort, 
Gloucestershire.  BMNH  D1795,  Inferior  Oolite,  ?locality.  BMNH  D 10091,  Pea  Grit  (murchisonae  Zone), 
Crickley  Hill,  Gloucestershire.  BMNH  D30002a-c,  Lower  Limestone  (murchisonae  Zone),  Kimsbury, 
Painswick,  Gloucestershire. 

Diagnosis.  Reptoclausa  with  continuous  autozooecial  ridges  separated  by  furrows  of  kenozooecia; 
zoaria  commonly  unilamellar,  occasionally  multilamellar. 

Description.  Zoaria  are  adnate,  fan-shaped  (PI.  88,  fig.  2)  to  discoidal,  commonly  unilamellar  but  occasionally 
multilamellar  (PI.  88,  fig.  3).  Zooecia  arise  where  existing  interzooecial  walls  divide  on  a basal  lamina  at  the 
colony  growth  margin.  Rounded  ridges  of  low  profile  cross  the  zoarial  surface  parallel  to  the  direction  of  growth 
and  form  lobate  projections  where  they  meet  the  colony  growth  margin  (PI.  88,  fig.  6).  Ridge  crests  are  about 
2 mm  apart  and  new  ridges  appear  at  dichotomies  of  established  ridges.  Ridges  are  occupied  by  autozooecia 
orientated  with  their  long  axes  slightly  divergent  from  the  ridge  crest.  Zooecium  size,  particularly  width, 
decreases  progressively  away  from  ridges  towards  intervening  furrows  occupied  by  kenozooecia  (text-fig.  2). 
Multilamellar  growth  was  achieved  either  by  spiral  overgrowth  around  irregularly  distributed  pivot  points 
(Taylor  1976),  or  by  a process,  comparable  with  the  frontal  budding  known  in  cheilostomes  (Banta  1972),  in 
which  an  overgrowing  zooecium  arose  from  an  autozooecial  aperture  to  initiate  a fan-shaped  expansion  on  the 
zoarial  surface.  The  first  zooecium  of  each  new  frontally  budded  layer  has  a short  frontal  wall  and  a 
longitudinally  elongate  aperture. 


text-fig.  2.  Semidiagrammatic  representation  of  part  of  a 
Reptoclausa  porcata  colony  showing  ridges  composed  of 
autozooecia  and  furrows  of  kenozooecia  lacking  aper- 
tures. Open  autozooecial  apertures  are  shown  in  black 
and  the  occluded  apertures  of  autozooecia  bordering 
kenozooecial  furrows  are  stippled.  The  large  zooecium  on 
the  left-hand  ridge  is  a gonozooecium.  The  lobate  distal 
growth  margin  is  evenly  stippled.  Approx,  x 18. 


704 


PALAEONTOLOGY,  VOLUME  23 


Frontal  walls  of  autozooecia  are  thick,  have  rounded  distal  terminations  (PI.  88,  fig.  5),  and  are  clearly  defined 
by  traces  of  vertical  interzooecial  walls  on  the  zoarial  surface,  Autozooecial  apertures  are  slightly  transversely 
elongate.  Thin-walled  peristomes  are  preserved  only  when  immured  by  intracolony  overgrowths.  Terminal 
diaphragms,  level  with  the  frontal  walls,  frequently  occlude  zooecia,  particularly  those  situated  at  boundaries 
between  ridges  and  furrows  (text-fig.  2).  Ontogenetic  zonation  (Silen  and  Harmelin  1974)  of  autozooecia  is  not 
apparent. 

Kenozooecia,  occurring  regularly  in  furrows  between  autozooecial  ridges  (PI.  88,  fig.  4),  have  narrow  frontal 
walls  defined  by  the  faint  traces  on  the  zoarial  surface  of  their  vertical  interzooecial  walls.  Less  elongate 
kenozooecia  may  occur  at  growth  margin  anastomoses  and  in  the  vicinity  of  zoarial  lateral  walls. 

Gonozooecia  are  developed  in  about  50%  of  the  zoaria  examined.  They  are  elongate,  slightly  dilated  in  width 
and  inflated,  and  are  situated  on  zoarial  ridges.  The  transversely  elongate  ooeciopores  are  about  the  same  size  as 
autozooecial  apertures. 

Dimensions.  See  Table  2. 


table  2.  Dimensions  (in  mm)  of  Reptoclausa  porcata  zooecia.  Abbrevia- 
tions as  in  Table  1. 


Nc 

Nz 

X 

Rc 

Rz 

law 

5 

125 

010 

009-0- 10 

0-07-0-11 

taw 

5 

125 

010 

0-09-0-11 

0-08-013 

fwl 

5 

125 

0-61 

0-52-0-66 

0-40-0-80 

fww 

5 

125 

0-22 

0-22-0-23 

0-18-0-29 

tgl 

4 

34 

1-70 

1-62-1-77 

1-17-2-25 

gw 

4 

36 

0-43 

0-37-0-46 

0-35-0-59 

low 

3 

19 

009 

0-08-0-09 

0-07-0-13 

tow 

3 

19 

0-12 

0-11-0-14 

0-10-0-15 

Remarks.  Among  the  specimens  included  by  Gregory  (18696)  in  Berenicea  allaudi  (Sauvage)  are  two 
(BMNH  D1794,  D1795)  belonging  to  this  new  species.  Rosacilla  allaudi  of  Sauvage  (1888)  is  a simple, 
multiserial  tubuloporinidean  lacking  ridged  zoaria  and  quite  distinct  from  the  species  figured  as 
Berenicea  allaudi  by  Gregory  (18966,  pi.  3,  fig.  6).  Walter  (1969,  p.  52)  includes  specimens  of 
Reptoclausa  porcata  within  Idmonea  triquetra  Lamouroux  1821.  Reptoclausa  porcata,  however, 
differs  from  Idmonea  triquetra  in  the  following  features: 

1 , R.  porcata  zoaria  have  a fan-shaped  to  discoidal  form  whereas  zoaria  of  /.  triquetra  consist  of  dichotomising 
narrow  multiserial  branches.  2,  The  branches  of  I.  triquetra  have  a well-defined  triangular  cross-section  distinct 
from  the  rounded  ridges  of  R.  porcata.  3,  Ooeciopores  of  I.  triquetra  are  about  half  the  diameter  of  R.  porcata 
ooeciopores.  4,  I.  triquetra  zooecia  are  usually  arranged  in  distinct  rows.  Those  of  R.  porcata  are  not  usually 
arranged  in  rows  and  have  larger  frontal  wall  dimensions. 

Furthermore,  R.  porcata  is  known  only  from  the  Upper  Aalenian  and  Lower  Bajocian,  whereas  the 
probable  range  of  I.  triquetra  is  Upper  Bajocian  to  Lower  Callovian  (Walter  1969). 

Hillmer  (1971,  p.  42)  noted  the  similarity  between  Lower  Cretaceous  Reptoclausa  and  two  of 
Walter’s  (1969)  figured  Idmonea  triquetra  specimens  (BMNH  D 10091,  D31586)  which  are  here 
included  in  R.  porcata.  R.  porcata  differs  from  the  Lower  Cretaceous  type-species  of  Reptoclausa, 
R.  neocomiensis  d’Orbigny  (redescribed  by  Hillmer  1971),  which  has  autozooecial  ridges  discon- 
tinuous in  the  direction  of  colony  growth  and  kenozooecia  occupying  a larger  proportion  of  the 
zoarial  surface.  The  known  range  of  the  genus  Reptoclausa  is  extended  back  from  the  Lower 
Cretaceous  into  the  Middle  Jurassic  by  the  description  of  R.  porcata. 

R.  porcata  is  abundant  in  the  Lower  Inferior  Oolite  of  the  Cotswolds  where,  along  with 
Reptomultisparsa  cricopora  and  R.  ventricosa,  it  is  found  encrusting  a variety  of  substrates  including 
the  large  terebratulid  brachiopod  Pseudoglossothyris  simplex  and  limestone  intraclasts.  Some  of  the 
brachiopod-encrusting  colonies  may  represent  associations  with  a living  brachiopod  because 


TAYLOR:  JURASSIC  BRYOZOA 


705 


bryozoan  growth  is  frequently  found  to  terminate  at  a growth  line  on  the  brachiopod  shell  suggesting 
that  growth  of  both  bryozoan  and  brachiopod  were  checked  simultaneously  but,  whereas 
brachiopod  growth  later  recommenced,  bryozoan  growth  did  not  (Ager  1961). 

Stratigraphical  range.  Upper  Aalenian-Lower  Bajocian. 

Discussion.  The  morphology  of  Reptoclausa  porcata  contrasts  with  that  of  most  other  multiserial 
encrusting  tubuloporinideans  (e.g.  Reptomultisparsa  tumida ) and  deserves  further  consideration.  In 
colonies  of  Reptoclausa  porcata  zooecium  size,  particularly  frontal  wall  width,  decreases  progres- 
sively passing  from  the  crests  of  the  ridges  to  the  bottoms  of  the  furrows  (text-fig.  2).  Ridge  crests 
bear  broad  autozooecia,  furrows  are  composed  of  narrow  kenozooecia,  and  the  intervening  regions 
between  ridge  crests  and  furrows  possess  comparatively  narrow  autozooecia  typically  occluded  by 
terminal  diaphragms.  This  type  of  morphological  gradient  perpendicular  to  the  growth  direction  of 
the  colony  suggests  the  presence  during  life  of  a physiological  gradient  (Bronstein  1939;  Anstey  et  at. 
1976),  perhaps  hormonal,  which  determined  zooid  structure  according  to  position  of  budding. 
Comparatively  large  zooids  were  budded  at  regularly  spaced  loci  along  the  growing  edge  of  the 
colony.  Here,  the  zoarium  was  differentially  thickened  to  give  a ridge  which  formed  a lobate 
projection  where  it  intersected  the  colony  growing  edge.  These  large  zooids  displayed  a dominance 
over  zooids  budded  between  loci  causing  them  to  be  crowded  and  reduced  in  size.  The  smallest  zooids 
budded  between  loci  were  too  small  to  support  a functional  polypide  (gut  and  tentacles)  and  thus 
became  kenozooids.  It  seems  that  zooids  of  intermediate  size  could  support  only  a short-lived 
polypide  whose  degeneration  was  followed  by  early  occlusion  of  the  zooecial  aperture  by  a terminal 
diaphragm.  The  functional  significance  of  the  unusual  arrangement  of  autozooecia  and  kenozooecia 
in  Reptoclausa  is  unclear  but  may  relate  to  the  maintenance  of  efficient  colony  feeding,  with 
autozooid  exhalent  currents  departing  from  the  colony  surface  above  regions  of  non-feeding 
kenozooids  (see  Taylor  1979). 

Acknowledgements.  I thank  Dr.  G.  P.  Larwood  (University  of  Durham)  for  constructive  criticism  of  the 
manuscript  and  for  his  supervision  during  the  course  of  a N.E.R.C.  funded  research  project.  Continued  support 
from  N.E.R.C.  in  the  form  of  a research  fellowship  is  also  gratefully  acknowledged.  Miss  J.  Darrell  kindly 
arranged  the  loan  of  specimens  from  the  British  Museum  (Natural  History).  Valuable  photographic  assistance 
was  provided  by  Dr.  J.  C.  W.  Cope  (University  College  of  Swansea). 


REFERENCES 

ager,  d.  v.  1961.  The  epifauna  of  a Devonian  Spiriferid.  Q.  Jl  geol.  Soc.  Lond.  117,  1 10. 
anstey,  R.  L.,  pachut,  J.  f.  and  prezbindowski,  d.  r.  1976.  Morphogenetic  gradients  in  Paleozoic  bryozoan 
colonies.  Paleobiology,  2,  131-146. 

banta,  w.  c.  1972.  The  body  wall  of  cheilostome  Bryozoa.  V.  Frontal  budding  in  Schizoporella  unicornis 
floridana.  Mar.  Biol.  14,  63-71. 

borg,  F.  1926.  Studies  on  Recent  cyclostomatous  Bryozoa.  Zool.  Bidr.  Upps.  10,  181-507. 
bronstein,  G.  1939.  Sur  les  gradients  physiologiques  dans  une  colonie  de  Bryozoaires.  C.r.  hebd.  Seanc.  Acad. 
Sci.,  Paris,  209,  602-603. 

busk,  G.  1852.  Catalogue  of  marine  Polyzoa  in  the  collection  of  the  British  Museum.  1.  Cheilostomata.  54  pp., 
British  Museum  (Natural  History),  London. 

canu,  f.  1918.  Les  ovicelles  des  Bryozoaires  Cyclostomes.  Etudes  sur  quelques  families  nouvelles  et  anciennes. 
Bull.  Soc.  geol.  Fr.  4th  series,  16,  324-335. 

ehrenberg,  c.  g.  1831.  Symbolae  physicae,  seu  leones  et  Descriptiones  Mammalium,  Avium,  Insectarum  et 
Animalium  Evertebratorum.  Pars  Zoologica,  Dec.  1 . 4.  Berlin. 

Gregory,  J.  w.  1894.  Catalogue  of  the  Jurassic  Bryozoa  in  the  York  Museum.  Rep.  Yorks,  phil.  Soc.  for  1893, 
58-61. 

— 1896a.  A revision  of  the  British  Jurassic  Bryozoa.  Part  III.  The  Genus  Berenicea.  Ann.  Mag.  nat.  Hist. 
6th  series,  17,  41-49. 

— 1896 b.  Catalogue  of  the  fossil  Bryozoa  in  the  Department  of  Geology,  British  Museum  ( Natural  History). 
The  Jurassic  Bryozoa.  239  pp.,  1 1 pis.,  British  Museum  (Natural  History),  London. 


706 


PALAEONTOLOGY,  VOLUME  23 


hillmer,  G.  1971.  Bryozoen  (Cyclostomata)  aus  dem  Unter-Hauterive  von  Nordwestdeutschland.  Mitt,  geol- 
palaont.  Inst.  Univ.  Hamburg , 40,  5-106. 

milne-edwards,  H.  1838.  Memoire  sur  les  Crisies,  les  Horneres  et  plusieurs  autres  Polypes  vivans  ou  fossiles 
dont  l’organisation  est  analogue  a celle  des  Tubulipores.  Annls  Sci.  nat.  Ser.  2,  9,  193-238,  pis.  6-16. 
orbigny,  a.  d\  1853.  Paleontologie  franqaise , terrains  cretaces,  5 Bryozoaires.  1,  1191  pp.,  2 (atlas),  pis.  600-800. 
Paris. 

pohowsky,  R.  A.  1978.  The  boring  ctenostomate  Bryozoa:  taxonomy  and  paleobiology  based  on  cavities  in 
calcareous  substrata.  Bull.  Am.  Paleont.  73,  No.  301,  1-192. 
sauvage,  h.  e.  1888.  Note  sur  les  Bryozoaires  Jurassiques  de  Boulogne.  Bull.  Soc.  geol.  Fr.  3rd  series,  27,  38-53. 
silen,  L.  and  harmelin,  J.-G.  1974.  Observations  on  living  Diastoporidae  (Bryozoa  Cyclostomata)  with  special 
regard  to  polymorphism.  Acta  zool.,  Stockh.  55,  81-96. 
taylor,  p.  d.  1976.  Multilamellar  growth  in  two  Jurassic  cyclostomatous  Bryozoa.  Palaeontology,  19, 293-306. 

— 1977.  The  palaeobiology  and  systematics  of  some  Jurassic  Bryozoa.  Ph.D.  thesis  (unpubl.),  University  of 
Durham  .518+ lviii  pp . 

— 1978.  A Jurassic  ctenostome  bryozoan  from  Yorkshire.  Proc.  Yorks,  geol.  Soc.  42,  211-216. 

— 1979.  The  inference  of  extrazooidal  feeding  currents  in  fossil  bryozoan  colonies.  Lethaia,  12,  47-56. 
thorpe,  j.  p.,  ryland,  j.  s.  and  beardmore,  j.  a.  1978.  Genetic  variation  and  biochemical  systematics  in  the 

marine  bryozoan  Alcyonidium  mytili.  Mar.  Biol.  49,  343-350. 
voigt,  E.  1977.  Arachnidium  jurassicum  n.  sp.  (Bryoz.  Ctenostomata)  aus  dem  mittleren  Dogger  von  Goslar  am 
Harz.  NeuesJb.  Geol.  Palaont.  Abh.  153,  170-179. 

Walter,  B.  1969.  Les  Bryozoaires  Jurassiques  en  France.  Etude  systematique.  Rapports  avec  la  stratigraphie  et 
la  paleoecologie.  Docums  Lab.  Geol.  Fac.  Sci.  Lyon,  35,  1-328,  pis.  1-20. 

PAUL  D.  TAYLOR 

Department  of  Palaeontology 
British  Museum  (Natural  History) 

Manuscript  received  2 October  1978  Cromwell  Road 

Revised  manuscript  received  20  June  1979  London,  SW7  5BD 


ANOMALOUS  OCCURRENCES  OF  THE  LOWER 
PALAEOZOIC  BRACHIOPOD  SCHIZOCRANIA 

by  m.  g.  lockley  and  D.  D.  j.  antia 


Abstract.  There  are  rare  occurrences  of  Ordovician  and  Silurian  species  of  the  inarticulate  brachiopod 
Schizocrania  attached  to  orthoconic  cephalopod  shells.  These  were  probably  transported  considerable  distances 
prior  to  their  deposition  in  onshore  sediments,  in  which  Schizocrania  is  not  normally  found.  Relationships 
between  host  and  encruster  are  discussed  with  a view  to  elucidating  both  encrustation  sequences  and  inferred 
ecological  associations. 


During  the  course  of  studies  of  Upper  Llanvirn,  Ordovician  (MGL)  and  Whitcliffian,  Silurian 
(DDJA)  successions  in  the  Anglo-Welsh  region,  we  noted  rare  occurrences  of  orthocones  with 
Schizocrania  (Trematidae)  attached  to  either  the  inner  or  outer  walls  of  their  body  chambers;  in  both 
cases  the  associated  clastic  sediments  are  of  a coarse  arenaceous  type  associated  with  demonstrably 
shallow-water  facies  (Williams  1953;  Antia  1979).  Havlicek  (1972,  p.  230)  reported  that  the  Upper 
Ordovician  trematid  Ptychopeltis  incola  Perner  from  Bohemia  \ . . lived  attached  only  to  the 
cylindrical  shells  of  orthocone  nautiloids’;  he  also  noted  that  its  ancestor  P.  hornyi  Havlicek 
sometimes  encrusted  orthocones.  We  therefore  consider  that  these  examples  of  apparent  host-specific 
relationships  may  be  paralleled  elsewhere  amongst  the  Trematidae  (e.g.  Schizocrania)  by  similar 
associations  between  host  and  encruster. 


MATERIAL 

The  Upper  Llanvirn  specimen  is  an  incomplete,  poorly  preserved  internal  mould  of  a body  chamber  of  an 
orthoconic  nautiloid  of  unknown  taxonomic  affinity.  It  was  recovered  from  a shell-bed  in  the  upper  part  of  the 
Flags  and  Grits  Formation  of  the  Ffairfach  Group  exposed  at  Coed  Duon,  3 km  south  of  Llangadog,  Dyfed 
(Grid  Ref.  SN  709256),  where  it  lay  parallel  to  bedding.  The  orthocone  has  three  specimens  of  Schizocrania  cf. 
salopiensis  Williams  attached  to  the  inner  surface  of  the  body  chamber;  the  brachial  valves  all  face  inwards  (text- 
fig.  1a)  but  show  no  obvious  preference  for  any  particular  attachment  site  although  two  of  the  specimens  are 
aligned  subparallel  to  each  other  near  the  anterior  end. 

The  Whitcliffian  specimens  are  represented  by  poorly  preserved  fragmentary  moulds  of  Orthoceras  sp. 
(diameters  c.  20  mm  and  > 30  mm  respectively)  from  the  Lower  Whitcliffe  Beds  of  Mortimer  Forest,  south  of 
Ludlow  (Grid  Ref.  SO  497725)  and  the  Upper  Whitcliffe  Beds  near  Broadstone  Farm  (SO  544900).  The  older 
specimen,  an  internal  mould  of  a large  portion  of  the  conch  (text-figs.  1b,  2b)  has  three  specimens  of  S.  striata 
(Sowerby)  attached  to  the  anterior  part  of  its  external  surface.  The  specimens  all  occur  close  to  each  other  on  the 
exposed  section  of  the  orthocone  mould  which  faces  downwards  from  the  undersurface  of  a bedded  unit;  relative 
to  the  final  entombment  position  of  the  orthocone  the  Schizocrania  specimens  occur  on  its  ‘underside’  and 
following  the  dissolution  of  the  cephalopod  shell  have  become  impressed  on  to  the  preserved  mould.  The 
younger  (upper  Whitcliffian)  specimen  consists  of  the  internal  and  external  moulds  of  a curved  fragment  of  a 
large  body  chamber;  it  has  five  poorly  preserved  specimens  of  S.  striata  attached  to  its  inner  (concave)  surface 
which  faces  downwards.  The  specimens  are  aligned  transversely,  parallel  to  the  peristome  (text-figs.  1,  2c). 

The  lectotype  (Geol.  Surv.  Mus.  No.  6631)  of  A.  striata  (Sowerby)  from  the  Leintwardinian-Whitcliffian  beds 
of  Delbury,  Salop  (Grid  Ref.  SO  501854)  is  the  only  other  known  British  Schizocrania  which  we  have  discovered 
attached  to  an  orthoconic  nautiloid  fragment;  the  specimen  is  attached  to  the  convex  surface  of  the  free  part  of  a 
septum,  probably  the  last  one;  it  differs  from  the  other  examples  in  its  larger  size  (length  9 mm)  and  posterior 
attachment  site  (text-fig.  2a). 


[Palaeontology,  Vol.  23,  Part  3,  1980,  pp.  707-713.1 


708 


PALAEONTOLOGY,  VOLUME  23 


■ection  of  for 

'ward  growth 

of 

Schizocrani 

>o  specimens 

length 

width(n 

BB  92492 

55 

60 

BB  92493 

42 

4-2 

BB 92494 

30 

30 

BB  94078 

3-7 

3-7 

BB  94079 

(3  6) 

3-6 

BB  94080 

2-3 

2-5 

BB  94081 

(40) 

(4-0) 

BB 94082 

3-5 

3 7 

BB  94083 

(4-0) 

(4-0) 

BB 94084 

(2-5) 

(2-5) 

BB 94085 

(2-5) 

(2-5) 

text-fig.  1.  Scale  drawings  of  Schizocrania  encrusted  orthoconic  nautiloids 
from  Upper  Llanvirn  strata  exposed  near  Llangadog,  Mid  Wales  (a)  and  from 
lower  (b)  and  upper  (c)  Whitcliffe  strata  exposed  near  Ludlow,  Salop.  All 
Schizocrania  specimens  have  British  Museum  numbers,  specimens  BB94078- 
94080  are  attached  to  the  outer  surface  of  the  shell  mould  (b)  whilst  the 
remaining  specimens  are  attached  to  the  inner  surfaces  of  the  shell  moulds. 

Approximate  length,  width  measurements,  listed  bottom  right. 


OBSERVATIONS  AND  INTERPRETATION 

All  twelve  of  these  Schizocrania  specimens  exhibit  only  their  convex  brachial  valves  facing  away  from 
the  cephalopod  shell  surface.  Schizocrania  is  known  to  attach  to  substrates  by  its  flat  pedicle  valve 
(Hall  and  Whitfield  1875;  Rowell  in  Williams  et  al.  1965,  p.  H283).  However,  pedicle  valves  are 
exceptionally  rare,  being  either  altogether  absent  from  assemblages  or  hidden  from  view  beneath  the 
brachial  valve.  The  three  orthocone  specimens  shown  in  text-fig.  1 indicate  that  the  anterior  edge  of 
the  phragmacone  was  the  preferred  encrustation  site  for  all  but  two  of  the  Schizocrania  specimens. 
The  orientation  of  these  Schizocrania  inside  the  phragmacone  and  on  the  shell  exterior  is  apparently 
not  random  since  all  adjacent  shells  are  aligned  with  their  umbones  pointing  in  approximately  the 
same  direction  (i.e.  transverse  or  oblique  to  the  orthocones’  long  axis). 


LOCKLEY  AND  ANTIA:  DISPERSAL  OF  SCHIZOCRANIA 


709 


The  orthocones  may  have  been  encrusted  while  they  were  alive  and  mobile,  or  when  they  were  dead 
and  floating,  or  dead  and  semi-buoyant,  being  washed  around  on  the  sea  floor,  or  dead  and  settled  on 
the  sea  floor,  or,  finally,  when  being  reworked. 

In  addition  to  the  numerous  examples  of  fossil  cephalopod  (ammonoid)  encrustation  recorded 
from  Mesozoic  assemblages  (e.g.  Seilacher  1960;  Meischner  1968)  and  the  few  broadly  analogous 
Lower  Palaeozoic  examples  involving  orthoconic  nautiloids  (Holland  1971;  Havlicek  1972),  we  have 
noted  Ordovician  and  Silurian  collections  containing  several  varied  and  undescribed  examples  of 
orthocone  encrustation  (e.g.  National  Museum  of  Wales  specimen  NMW  79.  5G.  Map  loc.  771; 
Hunterian  Museum  specimens  S. 25129/1-3  and  S. 251 14  a/b).  Schizocrania  is  ornamented  by 
numerous  radial  capillae  (Williams  1974,  p.  44).  According  to  Williams  and  Wright  1963,  p.  19  and 
Williams  and  Rowell  (in  Williams  etal.  1965,  p.  H81)  such  radial  ornament  probably  supported  setal 
follicles  at  the  commissure,  and  it  is  reasonable  to  assume  that  Schizocrania  was  particularly 
setiferous.  Sudarson  (1969,  p.  65)  noted  that  Discinisca  larvae  have  well-developed  principle  setae 
and  that  ‘there  may  be  a prolonged  larval  stage  . . . with  chaetae  increasing  in  number  to  facilitate 
floatation’.  Both  the  Schizocrania  species  discussed  here  exhibit  high  capillae  densities  at  the  same 


text-fig.  2.  a.  Schizocrania  striata  lectotype  showing  attachment  to  mould  of  orthocone  septum  from  upper 
Ludlow  beds,  Delbury,  Shrops.,  x 3.  B,  Detail  of  S.  striata  specimens  BB94078  {left)  and  BB94079  (right)  from 
Lower  Whitcliffe  Beds,  Mortimer  Forest,  Ludlow,  x 12-5;  see  also  text-fig.  1b.  c,  S.  striata  specimens  BB94081 
(top)  to  BB94085  showing  attachment  to  orthocone  body  chamber  fragment,  the  edge  of  which  is  arrowed,  from 
Upper  Whitcliffe  Beds,  Broadstone  farm,  Ludlow,  x 6.  Text-fig.  lc  is  a scale-drawing  of  the  counterpart  of  this 

specimen. 


710 


PALAEONTOLOGY,  VOLUME  23 


growth  stage  (i.e.  10-12  per  mm,  5 mm  antero-medianly  of  the  dorsal  umbones)  and  probably 
therefore  had  a juvenile  epiplanktic  stage. 

Holland  (1971,  p.  18)  considered  that  strophomenid  (aegeromenid)  and  rhynchonellid  ( Micro - 
sphaeridorhynchus  nucula ) brachiopods  might  have  attached  to  living  orthocone  hosts  but  concluded 
that  due  to  the  size  of  the  brachiopods  this  was  ‘unlikely’.  Havlicek  (1967,  p.  21)  demonstrated  the 
attachment  of  epiplanktic  strophomenids  to  the  ‘stems  of  algae’  (Havlicek  1967,  p.  21).  He 
subsequently  suggested  (Havlicek  1972,  p.  230)  that  aegeromenids  attached  to  live  orthocones  and 
considered  that  inarticulates  such  as  Ptychopeltis  incola  ‘were  attached  to  the  shells  of  living 
nautiloids’  (Havlicek  1972,  p.  230)  whilst  related  trematids  attached  both  to  orthocones  and  other 
specific  ‘freely  moving  organisms’  (Havlicek  1972,  p.  229).  An  orthocone  encrusted  with  Conchiolites 
(Ordovician)  was  described  by  Prantl  (1948,  p.  6).  Seilacher  (1954, 1968)  concluded  (1968,  p.  284)  that 
the  preferentially  orientated  epizoans  on  this  specimen  were  adjusted  to  the  ‘head-on  motion  of  their 
host’.  Both  Havlicek  (1972,  p.  230)  and  Seilacher  (1968)  suggested  that  preferred  orientation  of 
encrusters  is  of  prime  importance  in  testifying  to  pre-mortem  attachment.  This  suggests  that  the 
majority  of  known  Schizocrania  specimens  were  attached  at  various  stages  in  the  orthocone’s  post- 
mortem history.  Although  Havlicek  (1972,  p.  229)  presumed  that  aegeromenid  brachiopods  such  as 
those  depicted  by  Holland  (1971,  fig.  1 b)  attached  to  live  orthocones,  direct  evidence  for  this  is 
insubstantial.  Although  these  authors,  and  Bergstrom  (1968)  have  shown  such  brachiopods  attached 
in  rows  along  orthocones  and  ‘algal  stems’  such  arrangements  do  not  constitute  the  type  of  preferred 
orientation  referred  to  above. 

Since  modern  spirorbids  are  known  to  be  host  specific  and  capable  of  seeking  a preferred 
attachment  site  and  orientation  (Knight-Jones  1951),  it  is  almost  certain  that  the  occurrence  of  fossil 
spirorbids  aligned  along  the  growth  margins  of  orthocones  (Holland  1971)  indicates  a comparable 
relationship.  This  may  mean  that  the  similar  alignment  of  Schizocrania  specimens  noted  here  (text- 
fig.  1)  could  also  be  indicative  of  a host-specific  relationship.  Such  a contention  tends  to  be  supported 
by  our  observation  that  the  Anglo-Welsh  Schizocrania  have  not  been  found  attached  to  any  other 
host  organisms  and  would  also  offer  a possible  explanation  for  the  virtual  absence  of  pedicle  valves, 
which  could  have  either  remained  attached  to  a host  when  the  brachial  valve  disarticulated,  or 
become  obscured  during  fossilization  by  the  substrate  to  which  they  were  attached. 

The  Schizocrania  on  the  internal  surface  of  the  body  chambers  of  the  Llanvirn  and  upper 
Whitcliffian  specimens  indicate  encrustation  beginning  no  earlier  than  the  post-mortem  drifting 
phase  (following  decay  of  mantle  lining  the  body  chamber)  but  prior  to  the  infilling  of  the  body 
chamber.  Interpretation  of  the  lower  Whitcliffian  orthocones’  pre-entombment  history  is  prob- 
lematical; it  could  have  been  encrusted  at  any  one  of  a number  of  stages  in  its  history  as  a live  or  dead 
mobile  organism.  However,  since  the  Schizocrania  are  attached  to  its  ‘underside’  they  must  have 
settled  and  had  time  to  grow  prior  to  its  final  entombment  in  this  position.  The  S.  striata  lectotype 
must  have  become  attached  to  the  posterior  side  of  its  septal  substrate  after  the  separation  of  the 
orthocone’s  body  chamber  from  the  remaining  posterior  part  of  the  shell  (i.e.  at  a relatively  late  stage 
in  the  orthocones’  post-mortem  history). 

On  the  lower  Whitcliffian  orthocone  the  internal  mould  (text-fig.  1 b)  is  covered  by  numerous 
irregular  markings  consisting  mainly  of  small  elongated  raised  protruberances  averaging  about 
0T  mm  in  height  and  width  and  between  0-3  and  0-7  mm  in  length.  These  apparently  represent  the 
internal  moulds  of  bryozoan  borings  on  the  inner  surface  of  the  orthocone  shell  although  it  is  not 
altogether  clear  whether  some  of  the  flatter  or  even  slightly  indented  markings  may  not  result  from 
the  fossilization  of  external  borings.  In  any  event  where  the  Schizocrania  shells  are  slightly  broken, 
and  around  their  edges,  it  is  evident  that  the  borings  affect  the  orthocone  shell  beneath. 
Unfortunately  the  absence  of  a counterpart  of  this  specimen  renders  this  evidence  inconclusive. 

Distribution  of  Schizocrania 

The  Llanvirn  orthocone  and  Schizocrania  discussed  here  are  virtually  the  only  representatives  of 
these  taxa  known  from  the  predominantly  arenaceous  and  rudaceous  Ffairfach  Group  of  the 
Llandeilo  area.  Since  S.  salopiensis  is  common  in  penecontemporaneous,  argillaceous  successions 


LOCKLEY  AND  ANTIA:  DISPERSAL  OF  SCHIZOCRANIA 


711 


elsewhere  in  South  Wales  and  the  Welsh  Borderlands  (Williams  1974;  Bassett  et  al.  1974,  p.  9; 
Lockley  and  Williams,  in  press)  where  there  are  different  benthic  and  pelagic  faunas  (i.e.  trilobites, 
graptolites,  and  cephalopods),  it  is  reasonable  to  assume  that  the  exotic  Ffairfach  occurrence  may 
have  been  related  to  the  drifting  or  migration  of  a stray  cephalopod  beyond  the  normal  limits  of  its 
indigenous  environment.  Such  post-mortem  drifting  of  modern  cephalopods  is  well  known  (House 
1973;  Kennedy  and  Cobban  1976;  Hewitt  and  Pedlay  1978)  and  may  result  in  individual  specimens 
being  transported  for  hundreds  or  even  thousands  of  kilometres. 

Similarly  S.  striata  is  rare  in  the  Whitcliffe  Beds  of  the  Ludlow  region  where  it  constitutes  only 
about  0-01  to  0 005%  of  the  total  fauna  with  specimens  generally  occurring  in  a fragmentary 
condition  and  random  orientations.  It  is  more  common  in  unbioturbated,  parallel-laminated, 
alternating  light  and  dark  siltstones  (rhythmites)  of  deeper-water  facies  (e.g.  Upper  and  Lower 
Leintwardinian  Beds,  Holland  et  al.  1963,  p.  154;  Lawson  1973,  p.  274)  and  is  recorded  only  rarely  in 
shallow-water  bioturbated  siltstones  (Facies  B sensu  Antia  1979).  Again,  the  Whitcliffian  cephalo- 
pods drifted  into  inshore  deposits  from  an  offshore  source,  although  limited  evidence  also  points  to 
later  phases  of  encrustation  (e.g.  lectotype).  Williams  (1969,  p.  143)  discussed  the  potential  range  of 
larval  dispersal  and  its  bearing  on  brachiopod  migration  during  the  Ordovician.  Clearly  his  suggested 
figure  (up  to  250  km)  is  only  a fraction  of  the  range  potential  for  brachiopods  capable  of  encrusting 
live  or  drifting  orthocones. 

Trematid  hosts 

Encrusting  Trematidae  such  as  Schizocrania,  Drabodiscina,  and  Ptychopeltis  appear  to  have  been 
host  specific.  S.  salopiensis,  S.  striata,  and  P.  incola  have  hitherto  only  been  observed  attached  to 
orthoconic  nautiloids  generally  presumed  to  have  been  alive  or  floating  at  the  time  of  their 
encrustation.  Other  members  of  the  family,  e.g.  P.  hornyi  Havlicek  and  D.  grandis  Barrande,  are 
commonly  attached  to  conularids  which  are  considered  by  Havlicek  (1972)  to  have  been  mobile 
during  life,  and  the  American  species  S.  filosa  Hall  frequently  attached  to  the  brachiopod 
Rafinesquina  (e.g.  Cooper  1956  and  Rowell  in  Williams  et  al.  1965).  With  respect  to  trematid  - 
nautiloid  associations,  it  is  intriguing  to  note  that  Titus  and  Cameron  (1976)  record  S.  filosa  only  in 
their  deep-water  Geisonoceras  (Orthocerida)  community.  Dr.  R.  A.  Hewitt  and  Mrs.  D.  Evans  (pers. 
comm.  1979)  inform  us  that  they  know  of  no  Silurian  or  Ordovician  examples  of  cephalopod 
encrustation  by  brachiopods  other  than  those  reported  here,  which  is  suggestive  of  host-specific 
relationships. 


CONCLUSIONS 

Faunal  associations  with  abundant  Schizocrania  in  the  Ordovician  and  Silurian  of  the  Anglo-Welsh 
region  are  almost  invariably  confined  to  argillaceous  deep-water  facies  where  species  of  the  genus  are 
represented  almost  exclusively  by  assemblages  of  brachial  valves.  Such  exceptionally  dispropor- 
tionate valve  ratios  are  considered  to  result  from  their  encrustating  habits  which  might  account  for 
the  obscuring  or  removal  of  pedicle  valves.  Known  associations  between  trematid  encrusters  and 
hosts  such  as  those  reported  here  and  elsewhere  (e.g.  Havlicek  1972;  Rowell  in  Williams  et  al.  1965) 
point  to  some  form  of  host-specific  relationship  between  representatives  of  the  family  and  other 
larger  invertebrate  hosts.  Whether  such  relationships  could  be  termed  symbiotic,  commensal,  or 
parasitic  is  unclear  because  we  lack  evidence  which  demonstrates  that  hosts  were  encrusted  during 
life.  However,  we  can  establish  that  encrustation  of  orthocones,  which  may  in  some  cases  have  begun 
during  their  life,  often  began  no  earlier  than  the  post-mortem  drifting  phase,  and  may  have  continued 
or  begun  at  a time  when  the  orthocones  were  resting  or  rolling  on  the  sea  floor.  Since  encrustation  of 
many  of  these  orthocones  could  not  have  taken  place  when  they  were  in  the  final  ‘resting’  position  it 
must  have  occurred  during  the  middle  phases  of  their  pre-entombment  history. 

The  following  suggestions  on  the  time  of  encrustation  can  be  made:  (1)  The  encrusting 
Schizocrania  noted  here  are  not  currently  known  to  attach  to  non-orthocone  skeletal  components 
within  the  deposits  from  which  they  were  recovered  and  are  therefore  likely  to  have  settled 


712 


PALAEONTOLOGY,  VOLUME  23 


preferentially  on  orthocone  shells  prior  to  their  final  deposition.  (2)  The  apparent  high-density, 
orderly  clustering  of  Schizocrania  towards  the  anterior  of  the  conch  suggests  that  possibly  the 
orthocone  was  colonized  as  a specific  host  whilst  it  was  floating.  (3)  Since  both  Schizocrania  and  its 
nautiloid  hosts  are  normally  indigenous  to  sparsely  fossiliferous,  low-density  offshore  facies,  it  is 
probable  that  encrustation  occurred  in  an  offshore  region  before  the  orthocones  finally  became 
entombed  in  more  diverse,  fossiliferous,  onshore  facies  where  Schizocrania  is  invariably  rare.  This 
inference  is  supported  by  the  observation  that  the  setiferous  Schizocrania  may  well  have  been  adapted 
to  a prolonged  larval  stage  which  would  have  enhanced  its  chances  of  encountering  a suitable 
encrustation  site.  If  Schizocrania  even  occasionally  encrusted  orthocones  in  a manner  analogous  to 
the  attachment  of  epiplanktic  aegeromenids  to  buoyant  organisms  noted  by  Bergstrom  (1968),  then 
the  combined  effect  of  nautiloid  mobility  during  life  and  drifting  after  death  would  offer  an 
explanation  for  exceptionally  widespread  occurrences  of  certain  kinds  of  brachiopods. 

Acknowledgements.  We  thank  Dr.  G.  E.  Farrow  and  Dr.  R.  A.  Hewitt  for  critically  reading  the  manuscript;  Dr. 
M.  G.  Bassett,  Dr.  A.  Williams,  Dr.  J.  D.  Lawson,  Dr.  D.  Atkins,  Mrs.  D.  Evans,  Mr.  I.  Jarvis,  and  Dr. 
P.  Sheldon  are  also  thanked.  Both  authors  acknowledge  the  receipt  of  N.E.R.C.  grants. 


REFERENCES 

antia,  d.  d.  j.  1979.  Bone-Beds:  A review  of  their  classification,  occurrence,  genesis,  geochemistry,  ecology, 
diagenesis,  weathering  and  micro-biotas.  Mercian  Geol.  7,  93-174,  6 pis. 
bassett,  D.  a.,  ingham,  j.  K.  and  wright,  a.  d.  (eds.).  1974.  Field  Excursion  Guide  to  Type  and  Classical  Sections 
in  Britain.  (Ordovician  System  Symposium,  Birmingham  1974).  The  Palaeontological  Assoc.  London.  66  pp. 
bergstrom,  J.  1968.  Some  Ordovician  and  Silurian  Brachiopod  Assemblages.  Lethaia,  1,  23-237. 
cooper,  G.  a.  1956.  Chazyan  and  related  brachiopods.  Smithson.  Misc.  Coll.  127,  1-1245,  pis.  1-169. 
hall,  J.  and  whitfield,  r.  p.  1875.  Descriptions  of  invertebrate  fossils  mainly  from  the  Silurian  System.  Rep. 
Geol.  Surv.  Ohio , 2,  65-179. 

havlicek,  v.  1967.  Brachiopoda  of  the  Suborder  Strophomenida  in  Czechoslovakia.  Rozpr.  ustr.  Ust.  geol.  33, 
1-235. 

— 1972.  Life  habit  of  some  Ordovician  inarticulate  brachiopods.  Vestnik  ustred  ust.  geol.  47,  229-233. 
hewitt,  R.  a.  and  pedley,  H.  M.  1978.  The  preservation  of  the  shells  of  Sepia  in  the  Middle  Miocene  of  Malta. 

Proc.  Geol.  Ass.  89  (3),  227-237. 

Holland,  c.  H.  1971.  Some  conspicuous  participants  in  Palaeozoic  symbiosis.  Sci.  Proc.  R.  Soc.  Dublin,  Ser.  A, 
4,  15-26. 

— lawson,  j.  D.  and  walmsley,  v.  g.  1963.  The  Silurian  rocks  of  the  Ludlow  district,  Shropshire.  Bull.  Brit. 
Mus.  Nat.  Hist.,  Geol.  8,  95-171. 

house,  m.  r.  1973.  An  analysis  of  Devonian  Goniatite  distributions.  Spec.  Pap.  Palaeont.  12,  305-317. 
Kennedy,  w.  J.  and  cobban,  w.  D.  1976.  Aspects  of  Ammonite  Biology  and  Biostratigraphy.  Ibid.  17,  1-94. 
knight-jones,  e,  w.  1951.  Gregariousness  and  some  other  aspects  of  the  settling  behaviour  of  Spirorbis. 
J.  Marine  Biol.  Ass.  U.K.  30,  202-222. 

lawson,  J.  d.  1973.  Facies  and  faunal  changes  in  the  Ludlovian  rocks  of  Aymestry,  Herefordshire.  Geol.  J.  8, 
247-278. 

lockley,  m.  G.  and  williams,  A.  Lower  Ordovician  Brachiopoda  from  Mid  and  South  Wales.  Bull.  Br.  Mus.  Nat. 
Hist.  Geol.  (In  press.) 

meischner,  d.  1968.  Perniciose  Epokie  von  Placunopsis  auf  Ceratites.  Lethaia,  1,  156-174. 
prantl,  F.  1948.  The  genus  Conchicolites  Nicholson  (Serpulimorpha)  in  the  Ordovician  of  Bohemia.  Vestn. 
Krai.  Ceske  Spolecn.  Nauk.  (tr.)  9,  1 -7. 

seilacher,  a.  1954.  Okologie  der  triassichen  Muschel  Lima  lineata  (Schloth)  und  ihrer  epoken.  Neues  Jahrb. 
Geol.  Palaontol.  Monatsh.  4,  163-183. 

— 1960.  Epizoans  as  a key  to  Ammonoid  Ecology.  J.  Paleont.  34,  189-193. 

— 1968.  Swimming  habits  of  Belemnites— recorded  by  Boring  barnacles.  Palaeogeog.  Palaeoclimatol. 
Palaeoecol.  4,  279-285. 

sudarson,  A.  1969.  Brachiopod  larvae  from  the  west  coast  of  India.  Proc.  Ind.  Acad.  Sci.  68B,  59-68. 
titus,  r.  and  cameron,  b.  1976.  Fossil  Communities  of  the  Lower  Trenton  Group  (Middle  Ordovician)  of 
Central  and  North  Western  New  York  State.  J.  Paleont.  50,  1209-1225. 


LOCKLEY  AND  ANTIA:  DISPERSAL  OF  SCHIZOCRANIA 


713 


williams,  A.  1953.  The  geology  of  the  Llandeilo  district,  Carmarthenshire.  Q.  Jl  geol.  Soc.  Lond.  108,  177-208. 

— 1969.  Ordovician  faunal  provinces  with  reference  to  brachiopod  distribution.  In  wood,  a.  (ed.).  The  Pre- 
Cambrian  and  Lower  Palaeozoic  Rocks  of  Wales.  Univ.  of  Wales  Press,  Cardiff. 

— 1974.  Ordovician  Brachiopods  from  the  Shelve  district,  Shropshire.  Bull.  Brit.  Mus.  Nat.  Hist.  Geol.  Suppl. 
11,  1-163,  pis.  1-28. 

— and  wright,  A.  D.  1963.  The  Classification  of  the  ‘ Orthis  testudinaria  Dalman’  Group  of  Brachiopods. 
J.  Paleont.  37,  1-32. 

— et  al.  In  MOORE,  R.  c.  (ed.).  1965.  Treatise  on  Invertebrate  Paleontology.  Part  H.  Brachiopoda.  Univ.  Kansas 
Press. 


M.  G.  LOCKLEY 
Department  of  Geology 
University  of  Glasgow 
Glasgow,  G12  8QQ 
Scotland 


Manuscript  received  12  September  1979 
Revised  manuscript  received  12  December  1979 


D.  D.  J.  ANTIA 

B.P.  Development  Co.  Ltd. 
Fairburn  Estate 
Dyce 
Aberdeen 


THE  PALAEONTOLOGICAL  ASSOCIATION 


The  Association  was  founded  in  1957  to  further  the  study  of  palaeontology.  It  holds  meetings  and  demonstrations  as  well 
as  publishing  Palaeontology  and  Special  Papers  in  Palaeontology.  Membership  is  open  to  individuals  and  to  institutions  on 
payment  of  the  appropriate  annual  subscription.  Rates  for  1980  are: 

Institutional  membership  .....  £25  (U.S.  $60) 

Ordinary  membership  £12  (U.S.  $29) 

Student  membership  .....  £7-50  (U.S.  $18) 

There  is  no  admission  fee.  Correspondence  concerned  with  Institutional  Membership  should  be  addressed  to  Dr.  C.  H.  C. 
Brunton,  Department  of  Palaeontology,  British  Museum  (Natural  History),  Cromwell  Road,  London  SW7  5BD,  England. 
Student  members  are  persons  receiving  full-time  instruction  at  educational  institutions  recognized  by  the  Council.  On  first 
applying  for  membership,  an  application  form  should  be  obtained  from  the  Membership  Treasurer.  Subscriptions  cover  one 
calendar  year  and  are  due  each  January;  they  should  be  sent  to  the  Membership  Treasurer  (for  name  and  address  see 
Council  list  below). 


PALAEONTOLOGY 

All  members  who  join  for  1 980  will  receive  Volume  23,  Parts  1 -4.  All  back  numbers  are  still  in  print  and  may  be  ordered  from 
B.  H.  Blackwell,  Broad  Street,  Oxford  0X1  3BQ,  England,  at  £15  (U.S.  $36)  per  part  (post  free). 

SPECIAL  PAPERS  IN  PALAEONTOLOGY 

Members  may  subscribe  to  this  by  writing  to  the  Membership  Treasurer  (for  name  and  address  see  Council  list  below); 
the  subscription  rate  for  1980  is  £20  (U.S.  $48)  for  Institutional  Members,  and  £10  (U.S.  $24)  for  Ordinary  and  Student 
Members.  A single  copy  of  each  Special  Paper  is  available  to  Ordinary  and  Student  Members  only,  for  their  personal  use, 
at  a discount  of  25%  below  the  prices  shown  in  the  list  inside  the  front  cover.  Non-members  may  obtain  copies,  at  the 
listed  prices,  from  B.  H.  Blackwell,  Broad  Street,  Oxford  OX1  3BQ,  England. 


COUNCIL  1980-1981 

President:  Dr.  W.  H.  C.  Ramsbottom,  Institute  of  Geological  Sciences,  Leeds  LSI 5 8TQ 
Vice-Presidents:  Dr.  C.  T.  Scrutton,  Department  of  Geology,  The  University,  Newcastle  upon  Tyne  NE1  7RU 
Professor  C.  B.  Cox,  Department  of  Zoology,  King’s  College,  Strand,  London  WC2R  2LS 
Treasurer:  Mr.  R.  P.  Tripp,  High  Wood,  West  Kingsdown,  Sevenoaks,  Kent  TNI 5 6BN 
Membership  Treasurer  : Dr.  J.  C.  W.  Cope,  Department  of  Geology,  University  College,  Swansea  SA2  8PP 
Secretary:  Dr.  R.  Riding,  Department  of  Geology,  University  College,  Cardiff  CF1  1XL 


Editors 

Dr.  M.  G.  Bassett,  Department  of  Geology,  National  Museum  of  Wales,  Cardiff  CF1  3NP 
Dr.  K.  C.  Allen,  Department  of  Botany,  Bristol  University,  Bristol  BS8  1UG 
Dr.  R.  A.  Fortey,  Department  of  Palaeontology,  British  Museum  (Natural  History), 
Cromwell  Road,  London  SW7  5BD 

Dr.  A.  L.  Panchen,  Department  of  Zoology,  The  University,  Newcastle  upon  Tyne  NE1  7RU 


Other  Members  of  Council 


Dr.  R.  J.  Aldridge,  Nottingham 

Dr.  M.  D.  Brasier,  Hull 

Dr.  D.  E.  G.  Briggs,  London 

Dr.  C.  H.  C.  Brunton,  London 

Dr.  S.  Conway  Morris,  Open  University 

Dr.  R.  Harland,  Leeds 

Dr.  M.  B.  Hart,  Plymouth 


Dr.  P.  M.  Kier,  Washington 
Dr.  J.  Miller,  Edinburgh 
Dr.  T.  J.  Palmer,  Oxford 
Dr.  M.  Romano,  Sheffield 
Dr.  D.  J.  Siveter,  Leicester 
Dr.  N.  H.  Trewin,  Aberdeen 
Dr.  J.  Watson,  Manchester 


Overseas  Representatives 

Australia  : Professor  B.  D.  Webby,  Department  of  Geology,  Sydney  University,  Sydney,  N.S.W.,  2006 
Canada:  Dr.  B.  S.  Norford,  Institute  of  Sedimentary  and  Petroleum  Geology,  3303-33rd  Street  NW.,  Calgary,  Alberta 
New  Zealand:  Dr.  G.  R.  Stevens,  New  Zealand  Geological  Survey,  P.O.  Box  30368,  Lower  Hutt 
West  Indies  and  Central  America : Mr.  John  B.  Saunders,  Geological  Laboratory,  Texaco  Trinidad,  Inc.,  Pointe-a-Pierre, 
Trinidad,  West  Indies 

Western  U.S. A. : Professor  J.  Wyatt  Durham,  Department  of  Paleontology,  University  of  California,  Berkeley  4,  California 

Eastern  U.S.A.  : Professor  J.  W.  Wells,  Department  of  Geology,  Cornell  University,  Ithaca,  New  York 

South  America:  Dr.  O.  A.  Reig,  Departmento  de  Ecologia,  Universidad  Simon  Bolivar,  Caracas  108,  Venezuela 


Palaeontology 

VOLUME  23  ■ PART  3 

CONTENTS 

Dinoflagellate  cysts  from  the  Upper  Eocene-Lower  Oligocene 
of  the  Isle  of  Wight 

M.  LIENGJARERN,  L.  COSTA,  and  C.  DOWNIE 

475 

Dictyodora  from  the  Silurian  of  Peeblesshire,  Scotland 
M.  J.  BENTON  and  N.  H.  TRF.WIN 

501 

Lower  Cretaceous  Terebratulidae  from  south-western 
Morocco  and  their  biogeography 
F.  A.  MIDDLEMISS 

51.5 

Collignoniceratid  ammonites  from  the  Mid-Turonian  of 
England  and  northern  France 

W.  J.  KENNEDY,  C.  W.  WRIGHT,  and  J.  M.  HANCOCK 

557 

The  trilobite  Eccoptochile  from  the  Ordovician  of 
northern  Portugal 

M.  ROMANO 

605 

The  Miocene  horse  Hipparion  from  North  America  and 
from  the  type  locality  in  southern  France 
B.  J.  MACFADDEN 

617 

The  Toarcian  age  of  the  upper  part  of  the  Marlstone  Rock 
Bed  of  England 

M.  K.  HOWARTH 

637 

Jurassic  araucarian  cone  from  southern  England 
R.  A.  STOCKEY 

657 

Nomenclature  and  homology  in  peridinialean  dinoflagellate 
plate  patterns 
G.  L.  EATON 

667 

Mode  of  life  of  a giant  capulid  gastropod  from  the  Upper 
Cretaceous  of  Saghalien  and  Japan 

I.  HAYAMI  and  Y.  KANIE 

689 

Two  new  Jurassic  bryozoa  from  southern  England 

P.  D.  TAYLOR 

699 

Anomalous  occurrences  of  the  lower  Palaeozoic  brachiopod 
Schizocrania 

M.  G.  LOCKLEY  and  D.  D.  J.  ANTIA 

707 

Printed  in  Great  Britain  at  the  University  Press.  Oxford 
by  Eric  Buckley , Printer  to  the  University 

ISSN  0031-0239 

Palaeontology 

VOLUME  23  • PART  4 DECEMBER  1980 


Published  by 

The  Palaeontological  Association  London 


Price  £1  5 


THE  PALAEONTOLOGICAL  ASSOCIATION 

The  Association  publishes  Palaeontology  and  Special  Papers  in  Palaeontology.  Details  of  membership  and  subscription  rates 
may  be  found  inside  the  back  cover. 

The  journal  Palaeontology  is  devoted  to  the  publication  of  papers  on  all  aspects  of  palaeontology.  Review  articles  are 
particularly  welcome,  and  short  papers  can  often  be  published  rapidly.  A high  standard  of  illustration  is  a feature  of  the 
journal.  Four  parts  are  published  each  year  and  are  sent  free  to  all  members  of  the  Association.  Typescripts  on  all. aspects  of 
palaeontology  and  stratigraphical  palaeontology  are  invited.  They  should  conform  in  style  to  those  already  published  in  this 
journal,  and  should  be  sent  to  Dr.  R.  A.  Fortey,  Palaeontological  Association,  Department  of  Palaeontology,  British  Museum 
(Natural  History),  Cromwell  Road,  London  SW7  5BD,  England,  who  will  supply  detailed  instructions  for  authors  on  request 
(these  were  published  in  Palaeontology  1977,  20,  pp.  921-929). 

Special  Papers  in  Palaeontology  is  a series  of  substantial  separate  works;  the  following  are  available  (post  free). 

1.  (for  1967);  Miospores  in  the  Coal  Seams  of  the  Carboniferous  of  Great  Britain,  by  A.  H.  v.  smith  and  m.  a. 
butterworth.  324  pp.,  72  text-figs.,  27  plates.  Price  £8  (U.S.  $19.50). 

2.  (for  1968):  Evolution  of  the  Shell  Structure  of  Articulate  Brachiopods,  by  a.  williams.  55  pp.,  27  text-figs.,  24  plates. 
Price  £5  (U.S.  $12). 

3.  (for  1968):  Upper  Maestrichtian  Radiolaria  of  California,  by  Helen  p.  foreman.  82  pp.,  8 plates.  Price  £3  (U.S.  $7.50). 

4.  (for  1969):  Lower  Turonian  Ammonites  from  Israel,  by  R.  freund  and m.  raab.  83  pp.,  15  text-figs.,  10  plates.  Price  £3 
(U.S.  $7.50). 

5.  (for  1969):  Chitinozoa  from  the  Ordovician  Viola  and  Fernvale  Limestones  of  the  Arbuckle  Mountains,  Oklahoma, 
by  w.  a.  m.  jenkins.  44  pp.,  10  text-figs.,  9 plates.  Price  £2  (U.S.  $5). 

6.  (for  1969):  Ammonoidea  from  the  Mata  Series  (Santonian-Maastrichtian)  of  New  Zealand,  by  r.  a.  Henderson.  82  pp., 
13  text-figs.,  15  plates.  Price  £3  (U.S.  $7.50). 

7.  (for  1970):  Shell  Structure  of  the  Craniacea  and  other  Calcareous  Inarticulate  Brachiopoda,  by  A.  williams  and 
a.  D.  wright.  51  pp.,  17  text-figs.,  15  plates.  Price  £1  - 50  (U.S.  $4). 

8.  (for  1970):  Cenomanian  Ammonites  from  Southern  England,  by  w.  j.  Kennedy.  272  pp.,  5 tables , 64  plates.  Price  £8 
(U.S.  $19.50). 

9.  (for  1971):  Fish  from  the  Freshwater  Lower  Cretaceous  of  Victoria,  Australia,  with  Comments  on  the  Palaeo- 
environment,  by  m.  waldman.  130  pp.,  37  text-figs.,  18  plates.  Price  £5  (U.S.  $12). 

10.  (for  1971):  Upper  Cretaceous  Ostracoda  from  the  Carnarvon  Basin,  Western  Australia,  by  r.  h.  bate.  148  pp.,  43  text- 
figs. , 27  plates.  Price  £5  (U.S.  $12). 

11.  (for  1972):  Stromatolites  and  the  Biostratigraphy  of  the  Australian  Precambrian  and  Cambrian,  by  M.  R.  Walter. 
268  pp.,  55  text-figs.,  34  plates.  Price  £10  (U.S.  $24). 

12.  (for  1973):  Organisms  and  Continents  through  Time.  A Symposium  of  23  papers  edited  by  N.  F.  hughes.  340  pp., 
132  text-figs.  Price  £10  (U.S.  $24)  (published  with  the  Systematics  Association). 

13.  (for  1974):  Graptolite  studies  in  honour  of  O.  M.  B.  Bulman.  Edited  by  R.  b.  rickards,  d.  e.  jackson,  and c.  p.  hughes. 
261  pp.,  26  plates.  Price  £10  (U.S.  $24). 

14.  (for  1974):  Palaeogene  Foraminiferida  and  Palaeoecology,  Hampshire  and  Paris  Basins  and  the  English  Channel,  by 
j.  w.  Murray  and  c.  a.  wright.  171  pp.,  45  text-figs.,  20  plates.  Price  £8  (U.S.  $19.'50). 

15.  (for  1975):  Lower  and  Middle  Devonian  Conodonts  from  the  Broken  River  Embayment,  North  Queensland,  Australia, 
by  p.  G.  telford.  100  pp.,  9 text-figs.,  16  plates.  Price  £5-50  (U.S.  $13.50). 

16.  (for  1975):  The  Osfracod  Fauna  from  the  Santonian  Chalk  (Upper  Cretaceous)  of  Gingin,  Western  Australia,  by 
j.  w.  neale.  131  pp.,  40  text-figs.,  22  plates.  Price  £6-50  (U.S.  $16). 

17.  (for  1976):  Aspects  of  Ammonite  Biology,  Biogeography,  and  Biostratigraphy,  by  w.  J.  Kennedy  and  w.  a.  cobban. 
94  pp.,  24  text-figs.,  11  plates.  Price  £6  (U.S.  $14.50). 

18.  (for  1976):  Ostracoderm  Faunas  of  the  Delorme  and  Associated  Siluro-Devonian  Formations,  North  West  Territories, 
Canada,  by  D.  L.  dineley  and  e.  j.  loeffler.  218  pp.,  78  text-figs.,  33  plates.  Price  £20  (U.S.  $48). 

19.  (for  1977):  The  Palynology  of  Early  Tertiary  Sediments,  Ninetyeast  Ridge,  Indian  Ocean,  by  E.  m.  kemp  and  w.  K.  Harris. 
74  pp.,  2 text-figs.,  8 plates.  Price  £7  (U.S.  $17). 

20.  (for  1977):  Fossil  Priapulid  Worms,  by  S.  c.  morris.  159  pp.,  99  text-figs.,  30  plates.  Price  £16  (U.S.  $38.50). 

21.  (for  1978):  Devonian  Ammonoids  from  the  Appalachians  and  their  bearing  on  International  Zonation  and  Correla- 
tion, by  m.  r.  house.  70  pp.,  12  text-figs.,  10  plates.  Price  £12  (U.S.  $29). 

22.  (for  1978,  published  1979) : Curation  of  palaeontological  collections.  A joint  colloquium  of  The  Palaeontological  Associa- 
tion and  Geological  Curators’  Group.  Edited  by  m.  g.  bassett.  280  pp.,  53  text-figs.  Price  £25  (U.S.  $60). 

23.  (for  1979):  The  Devonian  System.  A Palaeontological  Association  International  Symposium.  Edited  by  M.  R.  house,  c.  t. 
scrutton,  and  m.  g.  bassett.  353  pp.,  102  text-figs.,  1 plate.  Price  £30  (U.S.  $72). 

24.  (for  1980):  Dinoflagellate  Cysts  and  Acritarchs  from  the  Eocene  of  southern  England,  by  J.  B.  bujak,  C.  downie,  g.  l. 
eaton,  and  G.  l.  williams.  104  pp.,  24  text-figs.,  22  plates.  Price  £15  (U.S.  $36). 

© The  Palaeontological  Association,  1980 


Cover:  Edriophus  levis  (Bather,  1914)  from  the  Middle  Ordovician  Trenton  Group  of  Kirkfield,  Ontario,  x 2-5. 
Specimen  in  the  Smithsonian  Institution;  photograph  by  H.  B.  Whittington. 


THE  TRILOBITE  TRETASPIS  FROM  THE  UPPER 
ORDOVICIAN  OF  THE  OSLO  REGION,  NORWAY 


by  ALAN  W.  OWEN 


Abstract.  All  known  Norwegian  species  of  Tretaspis  are  described.  Six  are  established  taxa:  T.  ceriodes 
(Angelin)  angelini  Stormer,  T.  seticornis  (Hisinger),  T.  anderssoni  Stormer,  T.  hadelandica  hadelandica  Stormer, 
T.  sortita  (Reed)  broeggeri  Stormer,  and  T.  kiaeri  Stormer.  Three  are  new:  T.  hisingeri,  T.  askerensis , and  T. 
latilimbus  (Linnarsson)  norvegicus.  Most  of  these  taxa  have  a broad  range  of  variation  encompassing  two  or 
more  morphs.  The  relative  proportions  of  these  morphs  are  used  to  distinguish  T.  latilimbus  norvegicus  and  T. 
sortita  broeggeri  from  their  nominate  subspecies.  The  British  form  T.  convergens  Dean  and  its  subspecies  are 
reinterpreted  as  subspecies  of  T.  hadelandica.  Ingham’s  concept  of  species  groups  within  Tretaspis  is  revised  with 
the  North  American  species  and,  provisionally,  T.  kiaeri  and  T.  calcaria  Dean  being  recognized  as  a distinct 
group  centred  on  T.  sagenosus  Whittington.  Neoteny  is  considered  to  have  played  a part  in  the  evolution  of 
Tretaspis. 

Species  of  the  trinucleid  Tretaspis  have  played  an  important  part  in  the  correlation  of  late  Caradoc 
and  Ashgill  successions  in  Britain  over  the  past  two  decades  (Dean  1961,  1963;  Ingham  1970;  Price 
1973,  1977;  McNamara  1979).  The  classical  studies  by  Stormer  (1930,  1945)  on  the  Scandinavian 
trinucleids  include  a number  of  species  of  Tretaspis , most  of  which  are  closely  related  to  British  forms. 
The  present  study  is  part  of  a broader  project  aimed  at  revising  the  late  Caradoc  and  Ashgill 
stratigraphy  and  trilobite  faunas  of  the  Oslo  Region.  This  area  was  divided  into  eleven  districts  by 
Stormer  (1953,  text-fig.  1)  and  Tretaspis  is  known  from  four  of  them  (text-fig.  1).  Of  these,  the  upper 
Ordovician  stratigraphy  of  two,  Hadeland  and  Ringerike,  has  been  revised  by  Owen  (1978, 1979)  and 
summaries  of  the  successions  in  the  other  two,  Oslo-Asker  and  Skien-Langesund,  are  given  by  Strand 
and  Henningsmoen  (1960,  pi.  7).  In  the  case  of  Oslo-Asker,  the  youngest  Ordovician  units  were 
redescribed  by  Brenchley  and  Newall  (1975).  The  present  study  includes  an  examination  of  all 
available  museum  material  and  samples  of  Tretaspis  collected  by  the  writer  from  1 1 7 localities  in 
Oslo-Asker,  Hadeland,  and  Ringerike,  now  housed  in  the  Paleontologisk  Museum,  Oslo  (PMO)  and 
the  Hunterian  Museum,  Glasgow  (HM). 

TREATMENT  OF  DATA 

The  distribution  of  pits  on  the  bilamellar  fringe  of  trinucleids  is  one  of  the  major  taxonomic  features 
of  the  group  (Hughes,  Ingham,  and  Addison,  1975,  pp.  550-552)  and  the  terminology  applied  herein 
is  that  advocated  by  Hughes  et  al.  (1975,  pp.  543-545,  text-figs.  3,  4).  Hughes  (1970)  demonstrated 
that  although  individual  specimens  of  Trinucleus  fimbriatus  Murchison  may  show  slight  asymmetry 
in  the  development  of  fringe  pits,  there  is  no  significant  statistical  difference  between  the  left  and  right 
sides  of  the  fringe  when  populations  are  considered.  It  has  thus  become  standard  practice  to  present 
data  in  terms  of  half-fringe  pit  counts  and  this  is  followed  herein.  Moreover,  Hughes  also 
demonstrated  that  the  distribution  and  number  of  pits  is  independent  of  holaspid  specimen  size  and 
this  also  is  assumed  for  other  trinucleids. 

The  number  of  arcs  of  pits  and  the  number  of  pits  in  each  arc  has  been  used  for  defining  species 
and  subspecies  in  various  trinucleids,  not  least  Tretaspis.  These  features  can  be  determined  even  in 
heavily  distorted  material  and  lend  themselves  to  simple  univariate  techniques  of  display  and 
analysis.  Such  an  approach  is  adopted  here  and  enables  direct  comparisons  to  be  made  with  data 
presented  in  other  studies.  Moreover,  few  horizons  in  the  Oslo  Region  have  yielded  more  than  a 

[Palaeontology,  Vol.  23,  Part  4,  1980,  pp.  715-747,  pis.  89-93.| 


716 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  1.  Stratigraphical  ranges  and  suggested  phylogeny  of  Norwegian  and  closely  related  British  species  of 
Tretaspis  in  terms  of  the  standard  British  succession.  The  geographical  distribution  of  Norwegian  forms  is  given 
also:  O-A  = Oslo-Asker  (Oslo  is  in  the  eastern  part  of  this  district),  H = Hadeland,  R = Ringerike,  S- 
L = Skien-Langesund.  The  British  species  are  revised  to  some  extent  herein. 

dozen  or  so  specimens  and  although  many  thousands  of  specimens  have  been  examined,  these 
comprise  relatively  few  complete  half-fringes,  let  alone  entire  fringes  and  thus  most  specimens  have 
provided  information  on  only  a small  proportion  of  the  possible  parameters.  Univariate  analysis 
therefore  is  preferred. 


PROBLEMS  OF  POLYMORPHISM 

Hughes  et  al.  (1975,  p.  590)  noted  that  in  many  trinucleid  stocks  there  is  a progressive  increase  in  the 
number  of  I arcs  between  1 1 and  In.  This  is  broadly  the  case  in  Tretaspis  and  in  general  terms  the  fringe 
criteria  used  in  defining  species  and  subspecies  are,  in  decreasing  order  of  importance:  (1)  the  number 
of  arcs  present,  (2)  whether  these  arcs  are  complete  anteriorly  and/or  posteriorly,  (3)  the  range  of 
variation  in  pit  number  per  arc  and  along  the  posterior  margin  of  the  fringe.  These  features  are  closely 
related  in  most  forms  in  that  there  is  a threshold  value  (4-7  pits  in  Norwegian  forms)  for  the  number 
of  pits  present  in  the  I arc  adjacent  to  In  before  that  arc  can  be  complete  anteriorly  and  a greater 
threshold  (7-13  pits  in  Norwegian  forms)  before  another  incomplete  arc  can  be  developed  between  it 
and  In.  Other  taxonomic  fringe  features  are  more  dependent  on  preservation  and  include  the  extent  of 
pits  in  adjacent  arcs  sharing  sulci,  the  size  of  individual  pits  and  the  development  of  lists  between  arcs. 

Many  of  the  species  and  subspecies  of  Tretaspis  described  from  Britain  appear  to  have  a fairly 
narrow  range  of  variation  with  a purely  typological  concept  based  on  characters  1 and  2 listed  above 
being  sufficiently  diagnostic  for  both  the  taxon  and  all  the  individuals  within  it.  In  some  cases  this 
may  be  simply  an  artefact  of  small  sample  sizes.  In  contrast,  Price  (1977,  pp.  764-772)  found  that 
some  populations  of  Tretaspis  from  Wales  have  a range  of  variation  in  fringe  characters  which 
encompasses  that  seen  in  two  named  taxa  which  he  considered  to  be  end-member  subspecies. 
Similarly,  Lesperance  and  Bertrand  (1976)  distinguished  a number  of  different  morphotypes  within 


OWEN:  TRILOBITE  TRETASPIS 


717 


Cryptolithus  although  this  needs  reassessment  in  relation  to  the  development  of  F pits  on  the 
posterior  part  of  the  fringe  (Owen  1980). 

Most  of  the  Norwegian  populations  of  Tretaspis  have  a broad  range  of  variation,  in  some  instances 
comprising  morphotypes  which  correspond  to  the  type  specimens  of  described  taxa.  This  presents 
great  problems  which,  if  taken  to  extremes  could  produce  a taxonomy  which  is  either  very  divisive 
and  unweildy  with  each  population  sample  comprising  a number  of  formally  named  taxa  or  one 
which  is  grossly  simplified  to  the  extent  of  masking  potentially  useful  affinities. 

A fairly  conservative  approach  is  therefore  adopted  with  different  phenotypes  within  populations 
being  recognized  by  the  neutral  term  ‘morphs’  (Mayr  1969,  p.  46).  In  the  case  of  T.  ceriodes  angelini 
there  is  evidence  for  the  progressive  establishment  of  distinct  phenotypes  (see  Hayami  and  Ozawa 
1975  for  a discussion  of  this  process).  In  all  the  other  Norwegian  taxa  only  slight  non-directional 
temporal  and  geographical  changes  in  the  relative  proportions  of  constituent  morphs  are  seen.  To 
some  extent  the  morphs  are  simply  the  product  of  variation  exceeding  the  threshold  values  noted 
above.  Thus,  for  example,  it  is  not  surprising  that  T.  ceriodes  angelini  morph  D (see  below),  which  has 
arc  I4  developed,  commonly  has  more  pits  in  I3  than  does  morph  B.  Moreover,  I3  is  always 
continuous  anteriorly  in  morph  D but,  by  definition,  is  incomplete  frontally  in  morph  B. 
Nevertheless,  the  recognition  of  morphs  is  found  to  be  very  useful  in  describing  variation  and  in 
making  comparisons  with  named  taxa  from  elsewhere  which  typologically  resemble  particular 
portions  of  the  Norwegian  range  of  variation. 

The  Norwegian  species  and  subspecies  therefore  are  defined  in  terms  of  recurrent  associations  of 
morphs  (Table  1)  and  to  some  extent  the  relative  percentages  of  these  morphs.  The  boundaries 
between  formally  named  taxa  in  some  instances  are  ones  of  convenience,  allowing  for  maximum 
stability  of  present  usage  within  the  new  framework.  Thus  as  Table  1 shows,  the  Norwegian  taxa 
T.  latilimbus  norvegicus  sp.  nov.  and  T.  sortita  broeggeri  are  distinguished  by  the  relative  abundance 
of  two  morphs  and  the  presence  in  the  latter  taxon  of  a third  morph.  This  allows  for  formal  expression 
of  the  greater  similarity  of  T.  latilimbus  norvegicus  to  the  Swedish  T.  latilimbus  latilimbus  (in  which 
morph  B is  virtually  absent)  and  of  T.  sortita  broeggeri  to  the  coeval  T.  sortita  sortita  from  Scotland 
which  is  composed  almost  entirely  of  morph  C. 


13 

14 

15 

Taxon 

Morph 

Radii 

E2 

*2 

% P. 

%c. 

%c. 

% p. 

%c. 

%c. 

%P 

%c. 

%c. 

°/ 

Pr.-.i 

Ant. 

Post. 

Ant 

Post. 

Ant. 

A 

34 

1 

c 

96 

8 

48 

T.  ceriodes  angelini 

B 

13 

1 

C 

c 

90 

0 

0 

C 

26 

| 

C 

c 

100 

18 

100 

D 

27 

1 

C 

c 

100 

66 

100 

100 

6 

14 

T.  hisingeri  sp.  nov 

2 

Inc 

•vine 

T.  seficornis 

2 

vine 

c 

T.  oskerensis  sp.  nov 

2 

Inc 

c 

29 

0 

0 

T.  anderssoni 

2 

Inc 

C 

100 

0 

0 

T.  hadelandica 

hade/andico 

A 

37 

2 

'■Jnc 

c 

100 

0 

61 

B 

56 

2 

;nc-C 

c 

100 

100 

23 

C 

7 

2 

Inc 

c 

ioo 

67 

100 

100 

o' 

0 

T.  latilimbus  norvegicus 

A 

59 

2 

c 

100 

100 

100 

100 

10 

29 

subsp.  nov 

B 

41 

2 

Inc 

c 

100 

100 

100 

100 

0 

17 

T.  sortita  broeggeri 

A 

5 

2 

c 

100 

100 

100 

100 

0 

0 

B 

58 

2 

Inc 

c 

100 

100 

100 

100 

0 

75 

C 

37 

2 

Inc 

c 

100 

100 

100 

100 

0 

93 

100 

ncT 

T.  kiaeri 

A 

65 

2 

c 

c 

100 

100 

100 

100 

3 

100 

B 

35 

2 

c 

c 

100 

100 

100 

100 

9 

100 

100 

0 

22 

table  1 . The  basic  fringe  development  in  the  Norwegian  species  and 
subspecies  of  Tretaspis.  Where  more  than  one  morph  is  recognized,  the 
pit  distribution  in  each  morph  is  given.  Note  that  the  range  in  pit 
number  in  each  arc  also  serves  to  differentiate  between  the  morphs  and 
this  is  detailed  in  text-figs.  2-9.  The  shading  marks  the  complete 
absence  of  an  arc,  Inc  = incomplete,  C = complete,  P = present. 
Post.  = posteriorly,  Ant.  = anteriorly 


718 


PALAEONTOLOGY,  VOLUME  23 


The  revision  of  the  Norwegian  Tretaspis  material  has  entailed  a reassessment  of  some  of  the  well- 
documented  British  forms.  In  addition,  Dr.  J.  K.  Ingham  of  Glasgow  University  has  given  me  access 
to  his  data  on  some  of  the  Swedish  forms.  Most  of  the  numerous  citations  in  the  literature  of  Swedish 
and  other  European  material  are  based  on  very  limited  collections,  as  is  the  North  American 
T.  clarkei.  The  polymorphic  Norwegian  material  indicates  that  individual  specimens  with  a 
particular  fringe  morphology  could  belong  to  one  of  a number  of  taxa.  Large  samples  are  required  to 
determine  the  range  of  variation  and  presence  of  morphs  before  taxonomic  assignment  can  be  carried 
out  with  any  confidence.  Thus  whilst  the  known  material  of  Tretaspis  from  outside  Norway  and 
Britain  is  discussed,  it  would  be  premature  to  make  more  than  general  comments  on  its  affinity. 


SYSTEMATIC  PALAEONTOLOGY 

Family  trinucleidae  Hawle  and  Corda,  1847 
Subfamily  trinucleinae  Hawle  and  Corda,  1847 
Genus  tretaspis  McCoy,  1849 

Type  species.  Asaphus  seticornis  Hisinger,  1 840,  p.  3,  pi.  37,  fig.  2;  from  the  Fjacka  Shale  (early  Ashgill),  Dalarna, 
Sweden;  by  subsequent  designation  of  Bassler  (1915,  p.  1285). 

Discussion.  Ingham  (1970,  pp.  41-45)  divided  Tretaspis  into  three  species  groups  centred  on 
T.  moeldenensis  Cave,  T.  seticornis  (Hisinger),  and  ‘7V  granulata  (Wahlenberg).  Hughes  et  al.  (1975, 
pp.  503-505)  reassigned  the  species  constituting  the  last-mentioned  group  to  Nankinolithus  Lu  and 
slightly  revised  the  other  two  groups. 

The  T.  seticornis  group  was  originally  stated  to  be  characterized  by  an  incomplete  or  absent  E2,  the 
1 1 — E j_2  radii  ‘out  of  phase’  with  those  containing  the  remaining  I arcs,  the  number  of  pits  in  Ex 
ranging  from  1 6 to  23,  rarely  up  to  27  (half-fringe),  the  thoracic  rachial  rings  relatively  broad  (tr.)  and 
bearing  a median  tubercle  and  the  pygidium  never  having  more  than  six  pairs  of  apodemes. 
Populations  described  below  as  T.  hadelandica  hadelandica  include  specimens  with  E2  complete  and 
up  to  ten  pairs  of  pygidial  apodemes.  Similarly,  populations  of  T.  anderssoni  have  seven  pairs  of 
apodemes.  In  all  other  respects  these  forms  correspond  to  the  T.  seticornis  group.  T.  persulcatus  from 
the  Upper  Drummuck  Group  at  Girvan,  south-west  Scotland,  has  a complete  E2  but  otherwise 
corresponds  to  the  T.  seticornis  group  and  was  almost  certainly  derived  from  an  unnamed  form  which 
has  E2  incomplete  (see  discussion  of  T.  hadelandica  below).  Thus  the  extent  of  E2  and  the  number  of 
pygidial  apodemes  are  not,  per  se,  indicative  of  the  T.  seticornis  group. 

Ingham  (1970,  pp.  44-45)  had  difficulty  in  assigning  T.  kiaeri  Stormer  to  his  groups  but  Hughes  et 
al.  (1975,  p.  563)  assigned  it  to  the  T.  moeldenensis  group.  T.  kiaeri  is  redescribed  here  and  has  E2 
complete  frontally,  two  sets  of  radii,  up  to  27^pits  in  Ex  and  up  to  ten  pairs  of  pygidial  apodemes.  It  is 
therefore  intermediate  between  the  T.  seticornis  and  the  T.  moeldenensis  groups.  T.  kiaeri  and  its 
probable  derivative  T.  calcaria  Dean  resemble  a number  of  middle  Ordovician  species  from  North 
America:  T.  canadensis  Staiible,  T.  reticulata  Ruedemann,  and  T.  sagenosus  Whittington  and  broadly 
coeval  allied  species  from  Scotland  and  Ireland.  (Hughes  et  al.  1975,  pp.  564-565).  These  middle 
Ordovician  forms  are  older  than  all  other  known  species  of  Tretaspis  and  have  a single  set  of  radii,  a 
large  number  of  pit  arcs  and  in  most  cases  a high  pit  count  in  most  arcs.  They  were  assigned  to  the  T. 
moeldenensis  group  by  Hughes  et  al.  (1975,  pp.  563-564).  Specimens  from  the  low  Carodoc  of 
Belgium  assigned  to  Tretaspis  by  Hughes  et  al.  (1975,  p.  564)  belong  to  Nankinolithus  (=  N.  sp.  of 
Hughes  et  al.  1975,  p.  559). 

The  T.  seticornis  group  as  presently  defined  seems  to  be  a natural  grouping  derived  in  the  earliest 
Ashgill  from  T.  ceriodes  (Angelin),  a member  of  the  T.  moeldenensis  group.  The  removal  of  T.  kiaeri, 
T.  calcaria,  and  the  middle  Ordovician  species  listed  above  would  leave  the  T.  moeldenensis  group  as  a 
close  grouping  within  which  phylogenetic  relationships  are  fairly  clear.  The  American  province  forms 
are  poorly  known  but  probably  closely  related  and  are  here  termed  the  T.  sagenosus  group.  They 
almost  certainly  gave  rise  to  T.  ceriodes,  the  earliest  known  member  of  the  revised  T.  moeldenensis 
group  possibly  by  neoteny  (giving  a much  simplified  fringe  morphology)  and  at  a time  of  major 


OWEN:  TRILOBITE  TRETASPIS 


719 


immigration  into  the  Scandinavian  area  (Bruton  and  Owen  1979).  T.  kiaeri  and  T.  calcaria  have  the 
typical  large  number  of  arcs  and  high  pit  counts  of  the  T.  sagenosus  group  and  whilst  having  two  sets 
of  radii  developed  the  possibility  exists  that  they  are  more  closely  related  to  that  group  than  to  the 
other  two  groups  and  thus  are  provisionally  included  in  it. 

Tretaspis  moeldenensis  group 
Tretaspis  ceriodes  ( Angelin,  1854)  angelini  Stormer,  1930 
Plate  89,  figs.  1-12;  text-fig.  2 

1887  Trinucleus;  Brogger,  p.  23. 

1930  Tretaspis  cerioides  [sic]  (Angelin);  Stormer,  pp.  44-48,  pi.  9,  figs.  1-4;  text-fig.  21  b. 

1930  Tretaspis  cerioides  var.  angelini  Stormer,  pp.  48-50,  pi.  9,  figs.  5-10. 

1934  Tretaspis  cerioides;  Stormer,  p.  331. 

1945  Tretaspis  ceriodes  (Angelin);  Stormer  (pars),  p.  402,  pi.  1 , fig.  6;  non  pp.  387, 404-405,  pi.  1 , fig.  7;  pi. 

4,  fig.  16  (=  T.  hadelandica  hadelandica). 

1945  Tretaspis  ceriodes  var.  angelini  Stormer;  Stormer,  p.  402,  pi.  1,  fig.  5. 

1945  Tretaspis  ceriodes  var.  donsi  Stormer,  pp.  388,  402,  405,  pi.  1,  fig.  8. 

1953  Tretaspis  ceriodes;  Stormer,  pp.  68,  87,  94. 

1953  T.  c.  angelini;  Stormer,  p.  68. 

1973  Tretaspis  cerioides;  Lauritzen,  p.  29. 

1978  Tretaspis  ceriodes  (sensu  lato)  (Angelin);  Owen,  pp.  9,  14,  15. 

1979  Tretaspis  ceriodes;  Owen,  pp.  250,  251. 

1979  Tretaspis  ceriodes  (Angelin)  ( sensu  lato);  Bruton  and  Owen,  text-figs.  3-6. 

Holotype.  A cranidium  (PMO  H226)  from  2 m below  the  top  of  the  Upper  Chasmops  Limestone  on 
Terneholmen,  Asker. 

Material,  localities,  and  horizons.  The  subspecies  has  a short  stratigraphical  range  and,  although  no  complete 
specimens  are  known,  a large  number  of  disarticulated  skeletal  elements  are  known  from  the  uppermost  parts  of 
the  Upper  Chasmops  Limestone  in  Baerum  and  Asker  in  the  western  part  of  Oslo- Asker  (see  Bruton  and  Owen 
1979  for  detailed  information),  from  0-85-1  02  m above  the  base  of  the  Lower  Tretaspis  Shale  on  Nakholmen, 
Oslo,  from  the  uppermost  parts  of  the  Solvang  Formation  throughout  Hadeland  and  at  Norderhov  in 
Ringerike,  and  from  the  lowest  part  of  the  Gagnum  Shale  Member  of  the  Lunner  Formation  in  the  northern  part 
of  Hadeland. 

Description.  Sagittal  length  of  glabella  equal  to  50-60%  of  width  between  posterior  fossulae.  Occipital  ring 
arched  gently  upwards  and  rearwards  and  defined  anteriorly  by  a shallow  furrow  which  bears  deep  slot-like  pits 
laterally.  Occiput  short  (sag.,  exsag.),  very  weakly  swollen.  Ip  furrows  deep,  transversely  oval.  2p  furrows  large, 
deep,  situated  a very  short  distance  in  front  of  lp  furrows  and  diverging  forwards  at  approximately  90°. 
Composite  lateral  glabellar  lobes  very  narrow  (tr.)  adjacent  to  2p  furrows,  anteriorly  and  posteriorly  to  which 
they  are  very  weakly  developed.  3p  furrows  developed  as  very  shallow  depressions  on  the  pseudofrontal  lobe 
directly  in  front  of  the  mid-length  of  the  glabella.  Pseudofrontal  lobe  very  strongly  swollen,  almost  circular  in 
dorsal  view,  occupying  approximately  70%  of  the  sagittal  glabellar  length.  Median  node  situated  on  the  highest 
part  of  the  glabella  at  60%  of  the  sagittal  glabellar  length.  Dorsal  furrows  broad  (tr.)  and  shallow  posteriorly, 
narrowing  and  deepening  a little  frontally,  diverging  forwards  at  approximately  30°  to  a level  a short  distance  in 
front  of  the  2p  furrows,  anteriorly  to  which  they  are  gently  convex  abaxially  and  bear  deep  fossulae  frontally. 
Genal  lobes  quadrant-shaped,  gently  inclined  from  the  dorsal  furrows,  more  steeply  declined  towards  the 
fringe.  Lateral  eye  tubercles  situated  opposite  or  slightly  in  front  of  2p  furrows.  Low  but  distinct  eye  ridges 
converge  adaxially  forwards  at  about  145°  from  the  eyes  to  the  outer  parts  of  the  dorsal  furrows.  Posterior 
border  furrows  deeply  incised,  transversely  directed,  bearing  deep  fossulae  distally.  Posterior  borders  ridge-like, 
transversely  directed  to  behind  posterior  fossulae  abaxially  to  which  they  are  deflected  steeply  downwards  and 
rearwards  at  approximately  60°.  External  surface  of  glabella  and  genal  lobes  bears  a variable  but  usually  strong 
reticulation  which  is  coarsest  around  the  glabellar  node  and  lateral  eye  tubercles.  On  internal  moulds  the  glabella 
is  commonly  smooth  and  the  genal  lobes  bear  a very  subdued  reticulation.  Fringe  flat-lying  over  the  inner  one  or 
two  I arcs  anteriorly  and  anterolaterally,  otherwise  almost  vertical. 

All  specimens  have  arcs  E1?  I x , I2,  and  In  complete  but  there  is  considerable  variation  in  the  development  of 
arcs  E2, 13,  and  I4.  On  the  basis  of  these  arcs,  four  morphs  are  recognized  (Table  1).  Morph  A lacks  E2  and  I4  and 


720 


PALAEONTOLOGY,  VOLUME  23 


has  I3  continuous  in  front  of  the  glabella  in  48%  of  3 1 specimens  where  this  could  be  determined  and  extending  to 
the  posterior  margin  in  8%  of  24  specimens.  I3  is  absent  in  4%  of  24  specimens.  Morph  B has  a complete  E2, 13 
developed  in  90%  of  20  specimens  but  never  continuous  anteriorly  or  posteriorly  and  I4  is  absent.  Morph  C has  a 
complete  E2,  no  I4,  and  an  I3  arc  which  is  always  continuous  anteriorly  and  extends  to  the  posterior  margin  in 
18%  of  36  specimens.  Morph  D has  E2  complete,  I3  invariably  complete  anteriorly  and  complete  posteriorly  in 
66%  of  25  specimens  and  a short  I4  developed.  The  range  of  variation  in  arcs  Ex,  I3,  In,  and  the  number  of  pits 
along  the  posterior  margin  of  the  fringe  for  T.  ceriodes  angelini  as  a whole  and  in  the  constituent  morphs,  is  given 
on  text-fig.  2 along  with  data  on  the  development  of  I4  in  morph  D.  With  the  exception  of  the  number  of  pits  in  I3, 
these  ranges  are  very  similar  for  all  the  morphs  although  the  mean  values  for  morph  A are  lower  than  those  of  the 
other  morphs.  Arcs  I1;  Ex,  and  E2  (when  present)  commonly  share  sulci  on  the  anterior  and  lateral  parts  of  the 
fringe  in  most  specimens.  Although  the  extent  of  this  feature  was  recorded  wherever  possible,  there  is  often  some 
difficulty  in  assessing  the  precise  extent  of  the  sulcation  which  may  also  be  partially  dependent  on  preservation 
and  consequently  this  is  not  presented  in  histogram  form.  In  a few  specimens  the  sulcation  does  not  extend 
laterally  beyond  the  dorsal  furrows  and  in  a few  it  extends  almost  to  the  posterior  margin.  The  mean  extent  is  to 
about  bR9  (fifty  specimens,  standard  deviation  4)  and  there  is  no  apparent  difference  between  the  morphs.  Only 
one  set  of  radii  is  developed.  On  external  surfaces,  lists  are  developed  between  all  the  I arcs.  Genal  spines  parallel, 
length  unknown. 

Hypostoma  and  thorax  not  known. 

Pygidium  sub-semicircular  in  outline  with  sagittal  length  equal  to  approximately  35%  of  the  anterior  width. 
Rachis  occupies  25%  of  the  anterior  width  of  the  pygidium,  tapers  rearwards  at  about  30°,  and  is  composed  of  an 
anterior  articulating  half-ring  and  five  or  six  rings.  Ring  furrows  progressively  less  well-defined  rearwards  along 
the  rachis,  bearing  deep  apodemal  pits  a short  distance  in  from  the  weakly  incised  dorsal  furrows.  Pleural  lobes 
flat-lying,  bearing  four  pairs  of  very  broad  furrows  which  define  three  or  four  ribs  which  die  out  some  distance 
from  the  weakly  developed  marginal  rim.  Pygidial  border  very  steeply  declined,  broad,  maintaining  constant 
width.  Antero-lateral  corners  of  pygidium  bear  steeply  declined  facets  which  diverge  abaxially  backwards  at 
about  120°. 

Discussion.  The  absence  of  arc  E2  from  morph  A clearly  distinguishes  it  from  the  other  morphs  where 
this  arc  is  not  only  present,  but  complete.  Morphs  B,  C,  and  D could  be  viewed  as  representing  a 
single  morphological  type  with  a broad  range  of  variation.  However,  three  morphs  are  recognized 
because  two,  B and  C,  are  similar  to,  or  correspond  to,  the  holotypes  of  named  taxa,  and  there  is  also 
some  evidence  for  a progressive  development  of  levels  of  phenotype  organization  from  morph  A 
through  B and  C to  D.  In  Hadeland,  a sample  of  thirty-three  specimens  from  an  exposure  of  the 
Lieker  Member  of  the  Solvang  Formation  illustrated  by  Owen  (1978,  text-fig.  6)  from  a level  near  the 
first  appearance  of  the  species  has  the  following  morph  composition:  A88%,  B6%,  and  C6%.  Higher 
levels  in  the  formation  in  the  nearby  stratotype  section  have  yielded  morph  D,  and  ten  specimens 
from  broadly  equivalent  levels  in  the  Gagnum  Shale  (including  the  holotype  of  T.  ceriodes  donsi ) 
comprise  B10%,  C80%,  and  D10%.  Similar  results  have  been  obtained  from  Oslo-Asker  with  early 
populations  having  morph  A dominant  over  B and  C;  morph  D being  restricted  to  the  later 
populations  where  A is  rare  or  absent. 

It  can  be  argued,  therefore,  that  morph  A represents  the  primitive  condition,  the  development 
of  a complete  E2  arc  in  some  members  of  the  population  giving  morphs  B and  C and  individuals  of 
morph  D type  developed  from  morph  C parents.  It  must  be  stressed,  however,  that  the  morphs 
are  regarded  as  representing  fairly  broad  portions  of  the  range  of  variation  in  interbreeding 
populations. 

Angelin’s  original  material  of  T.  ceriodes  (1854,  p.  65,  pi.  34,  fig.  2-2 b)  from  the  Upper  Mossen 
Formation  (late  Caradoc)  at  Kinnekulle,  Vastergotland,  Sweden,  was  reported  by  Stormer  (1930,  p. 
45)  to  be  lost  and  a neotype  from  the  Solvang  Formation  in  Ringerike  was  chosen.  This  neotype  could 
not  have  any  standing  as  it  was  not  from  the  type  locality  and  recently  Angelin’s  probable  syntypes 
have  come  to  light  in  the  collections  of  the  Riksmuseum,  Stockholm.  A full  examination  of  the  E pit 
development  can  be  made  in  only  one  of  these  and  E2  is  not  developed.  Two  specimens  show  the 
development  of  I3  which  in  both  cases  is  short  (3-4  pits)  and  not  present  anteriorly.  I4  is  absent.  Thus 
these  probable  syntypes  resemble  T.  ceriodes  angelini  morph  A.  Two  other  specimens  in  the 
Riksmuseum  collections  from  the  Upper  Mossen  Formation  (locality  not  known)  show  an  extensive 
I3  development  and  while  one  lacks  E2,  the  other  has  it  developed  mesially  but  not  beyond  R4.  This 


OWEN:  TRILOBITE  TRETASPIS 


D n = l8 
C n = 21 
B n = l3 


A I3  complete  anteriorly  n=8 
A I3  incomplete  anteriorly  n=9 
A all  specimens  n=20 


A 13  incomp.  ant.  1 — M ^ n = 10 
A all  specimens  1 — mm  M n=25 


text-fig.  2.  Histograms  showing  the  range  of  variation  in  fringe  characters  of  all 
available  specimens  of  Tretaspis  ceriodes  angelini  with  a comparison  of  the 
range,  mean,  and  one  sample  standard  deviation  on  each  side  of  the  mean  of  the 
four  morphs  (A,  B,  C,  and  D)  present  in  the  subspecies.  Morph  A is  also 
subdivided  to  compare  these  parameters  in  specimens  with  arc  I3  incomplete 
anteriorly  (i.e.  like  morph  B)  with  those  in  which  this  arc  is  complete  anteriorly 
(i.e.  like  morphs  C and  D).  It  may  prove  useful  to  define  separate  morphs  on  this 
basis  once  more  material  is  available.  In  all  instances  n = number  of  specimens 
in  the  sample. 


722 


PALAEONTOLOGY,  VOLUME  23 


condition  is  not  known  from  any  Norwegian  specimen.  Detailed  comparisons  of  the  Swedish  and 
Norwegian  forms  must  await  the  documentation  of  more  material  from  Kinnekulle. 

T.  ceriodes  alyta  Ingham,  1970,  from  the  upper  part  of  the  Onnian  Stage  in  northern  England  has 
arcs  E2  complete,  I4  absent,  and  I3  extensive  or  complete  posteriorly  but  incomplete  anteriorly.  It 
thus  resembles  T.  ceriodes  angelini  morph  B,  differing  only  in  having  a more  extensive  I3  arc  and  the 
I1-E1_2  sulci  commonly  extending  almost  to  the  genal  angles.  Examination  of  specimens  from  the 
Onnian  Stage  in  the  Gross  Fell  Inlier  in  northern  England  figured  by  Dean  (1961,  1962)  shows  that 
Ingham  was  correct  in  suggesting  that  they  belong  to  T.  ceriodes  alyta  (1970,  p.  5).  Some  of  the 
specimens  of  supposed  Onnian  age  in  Dean’s  collections  in  the  Cross  Fell  Inlier  (localities  A12  and 
A1 5 of  Dean,  1959,  text-fig.  1)  have  a very  large  number  of  pits  in  In  (25^-28^)  and  up  to  9\  pits  in  I4 
and  most  closely  resemble  T.  moeldenensis  Cave,  1960  (see  Price  1977,  pp.  764-772  for  a discussion  of 
this  species). 

T.  ceriodes  favus  Dean,  1963,  is  a poorly  known  form  based  on  specimens  from  the  upper  part  of 
the  Actonian  Stage  and  the  lowest  beds  of  the  Onnian  Stage  in  the  Onny  River  section  and  supposed 
Actonian  strata  near  Cardington,  Salop,  England.  The  subspecies  was  diagnosed  as  having  arc  E2 
developed  only  laterally  and  I3  complete  anteriorly  but  not  posteriorly.  Whilst  the  latter  is  true  for  the 
holotype  and  other  specimens  from  the  Onny  River,  the  material  is  too  poorly  preserved  for  the  E pit 
development  to  be  discerned  fully  although  E2  is  certainly  present.  The  I3  development  is  closest  to 
that  seen  in  T.  ceriodes  angelini  morph  C.  All  of  the  sixteen  specimens  from  near  Cardington  in  the 
British  Museum  (Natural  History)  (including  Dean  collection)  and  the  Hunterian  Museum  (Owen 
and  Ingham  collection)  in  which  the  E arc  development  is  clear,  undoubtedly  have  E2  complete.  I3  is 
incomplete  anteriorly  in  this  material  (eleven  specimens)  and  has  2-14  pits.  Arcs  \x-YLl_2  are  sulcate 
over  almost  the  whole  fringe.  The  Cardington  material  therefore  is  similar  to  both  T.  ceriodes  angelini 
morph  B and  T.  ceriodes  alyta  and,  as  noted  by  Bruton  and  Owen  (1979,  p.  220),  its  association  with 
Onnia  gracilis  may  indicate  an  Onnian  age  for  the  strata  here. 

T.  ceryx  Lamont,  1941,  from  the  Raheen  Shales  (late  Caradoc-early  Ashgill)  of  Co.  Waterford, 
Eire,  differs  from  T.  ceriodes  angelini  morph  C only  in  having  very  long,  slot-like  I1-E1_2  sulci 
anteriorly  and  anterolaterally.  The  Irish  form  is  probably  best  viewed  as  a geographical  subspecies  of 
T.  ceriodes. 

T.  colliquia  Ingham,  1970,  from  the  Pusgillian  Stage  in  the  Cautley  district  of  northern  England  is 
probably  a derivative  of  T.  ceriodes  alyta  and  some  specimens,  like  T.  ceriodes  angelini  morph  D have 
a short  I4  developed.  The  English  species  is  distinguished  by  its  very  large,  deep,  extensive  I1-E1_2 
sulci  and  in  having  a very  high  E pit  count  (twenty-eight  in  the  two  specimens  showing  this  feature). 

Dr.  J.  K.  Ingham  of  Glasgow  University  has  informed  me  of  an  undescribed  form  of  T.  ceriodes 
similar  to  T.  ceriodes  angelini  morph  A from  the  Upper  Whitehouse  Group  (late  Caradoc-early 
Ashgill)  at  Girvan,  south-west  Scotland  (Ingham  1978,  pp.  170,  171). 


EXPLANATION  OF  PLATE  89 

Figs.  1-12.  Tretaspis  ceriodes  (Angelin)  angelini  Stormer.  1,  3,  5,  morph  D,  PM0100826,  dorsal,  anterior,  and 
lateral  views  of  internal  mould  of  cranidium,  5-3-54  m below  top  of  Solvang  Formation,  Norderhov, 
Ringerike,  x 4.  2,  4,  morph  D,  PMO101552,  dorsal  and  anterolateral  views  of  external  surface  of  cephalon, 
approximately  1-7  m below  top  of  Upper  Chasmops  Limestone,  East  Raudskjer,  Asker,  x 6.  6,  holotype, 
morph  C,  PMO  H226,  oblique  anterolateral  view  of  partially  exfoliated  cranidium,  2 m below  top  of  Upper 
Chasmops  Limestone,  Terneholmen,  Asker,  x 6^;  also  figured  by  Stormer  (1930,  pi.  9,  fig.  5).  7, 10,  morph  A, 
PMO  H593,  posterolateral  and  frontal  views  of  partially  exfoliated  cephalon,  same  horizon  and  locality  as  6, 
x 5;  also  figured  by  Stormer  (1930,  pi.  9,  fig.  10).  8,  morph  B,  PMO  H250,  anterolateral  view  of  partially 
exfoliated  cephalon,  same  horizon  and  locality  as  6,  x 10.  9,PMO103952,dorsalviewofpygidium,upperpart 
of  Solvang  Formation,  Lunner,  Hadeland,  x 4J.  11,  morph  C,  PM081 100,  anterolateral  view  of  partially 
exfoliated  small  cranidium,  same  horizon  and  locality  as  2,  x 20.  1 2,  morph  A,  PMO  H495,  anterolateral  view 
of  partially  exfoliated  cephalon,  0-85-1-02  m above  base  of  Lower  Tretaspis  Shale,  Nakholmen,  Oslo,  x 4. 


PLATE  89 


B I 


v m 


OWEN,  trilobite  Tretaspis 


724 


PALAEONTOLOGY,  VOLUME  23 


Tretaspis  seticomis  group 
Tretaspis  seticornis  (Hisinger,  1840) 

Plate  90,  figs.  1 -4 

1840  Asaphus  seticomis  Hisinger,  p.  3,  pi.  37,  fig.  2. 

1840  Asaphus  cyllarus  Hisinger,  p.  3,  pi.  37,  fig.  3. 

71845  Trinucleus  seticomis  (Hisinger);  Loven,  p.  107,  pi.  2,  fig.  2. 

71854  Trinucleus  seticornis  (Hisinger);  Angelin,  p.  84,  pi.  40,  fig.  19. 

71869  Trinucleus  seticornis  (Hisinger);  Linnarsson,  p.  79. 

1883  Trinucleus  seticornis  (Hisinger);  Tornquist,  p.  43. 

71884  Trinucleus  seticornis  (Hisinger);  Tornquist,  pp.  84-87. 

71887  Trinucleus  seticornis  (Hisinger);  Brogger,  p.  24. 

1930  Tretaspis  seticornis  (Hisinger);  Stormer  (pars),  pp.  55-67,  ?pl.  7;  ?pl.  8;  ?pl.  1 1,  fig.  4;  text-figs.  27, 28 
(pars),  729,  33a,  346  (pars),  34c,  736,  ?37a,  b,  742. 

1934  Tretaspis  seticornis',  Stormer  (pars),  p.  330. 

1936  Tretaspis  seticornis  (Hisinger);  Asklund  (pars),  p.  4,  pi.  1,  figs.  1-3,  75,  76,  non  4. 

71959  Tretaspis  seticornis  (Hisinger);  Whittington  in  Moore,  text-fig.  323.2. 

1979  Tretaspis  seticornis  seticornis  (Hisinger);  Owen,  pp.  250,  251,  252,  text-fig.  6. 

1979  Tretaspis  seticomis  seticornis  (Hisinger);  Bruton  and  Owen,  text-fig.  6. 

This  synonomy  only  includes  references  to  material  which  actually,  or  very  probably,  belongs  to  T.  seticornis.  A 
more  complete  list,  comprising  forty-seven  entries,  was  given  by  the  writer  (1977,  pp.  243-245)  in  an  unpublished 
thesis  and  includes  reidentifications  wherever  possible. 

Material,  localities,  and  horizons.  Hisinger’s  syntypes  of  Asaphus  seticornis  from  the  Fjacka  Shale  in  well  diggings 
at  Furudal  in  Dalarna,  Sweden,  have  not  been  identified  unequivocally  in  the  collections  of  the  Riksmuseum, 
Stockholm,  and  as  noted  by  Tornquist  (1883,  p.  43)  may  not  have  been  collected  in  situ.  The  species,  as  here 
defined,  is  known  from  the  lower  part  of  the  Fjacka  Shale  (J.  K.  Ingham,  pers.  comm.  1 976),  the  lower  part  of  the 
Lower  Tretaspis  Shale  at  Ole  Deviks  Vei  (lowest  5-86  m),  Astaddammen  (lowest  4-65  m at  least),  S.  Grakommen 
and  between  Fossung  and  Hogstad  in  Oslo-Asker,  and  from  the  Hogberg  Member  of  the  Solvang  Formation  on 
Frognoya,  Ringerike. 

Description.  Most  of  the  available  material  is  crushed  to  some  extent.  Glabella  and  genal  lobes  similar  to  those  of 
T.  ceriodes  angelini  except  that  the  pseudofrontal  lobe  is  more  elongate.  External  surface  of  glabella  and  genal 
lobes  smooth  or  bearing  a faint  reticulation.  Internal  moulds  smooth.  Steeply  declined  fringe  bears  complete  arcs 
Els  Ix,  I2,  and  In,  and  an  incomplete  E2  arc.  Arcs  Ix  -Ex_2  are  out  of  phase  with  radii  comprising  the  other  two  I 
arcs.  Pits  in  Ix  and  Ex  share  sulci  anteriorly  and  anterolaterally.  There  is  insufficient  material  to  assess  the  range 
of  variation  in  pit  distribution.  Only  one  Norwegian  specimen  is  sufficiently  well  preserved  for  the  number  in  Ex 
to  be  determined  (18),  and  whilst  one  specimen  clearly  lacks  E2,  others  show  minimum  values  of  4,  7,  9 (3 
specimens),  and  1 0 pits.  In  is  seen  completely  in  4 specimens  where  it  comprises  1 5, 17),  18,  and  1 8)  pits  and  there 
are  6 (3  specimens)  or  7 (3  specimens)  pits  along  the  posterior  margin  of  the  fringe.  Lists  are  not  developed.  One 
specimen  (pi.  90,  fig.  4)  does  not  conform  to  the  typical  T.  seticornis  development  in  having  a stronger 
reticulation  and  in  having  pits  developed  in  I3  on  the  lateral  parts  of  the  fringe  at  aR6,  7,  9,  11-17.  Such  a 
development  is  most  unusual  for  any  species  of  Tretaspis  and  may  reflect  hybridization  with  T.  hadelandica 
hadelandica  which  includes  morphs  with  this  arc  complete  posteriorly. 

Hypostoma  unknown. 

Thorax  barrel-shaped,  comprising  six  segments  of  which  the  third  and  fourth  are  slightly  broader  (tr.)  than  the 
rest.  Rachis  occupies  30%  of  the  width  of  each  segment  and  is  bounded  laterally  by  very  weakly  incised  dorsal 
furrows.  Rachial  rings  strongly  convex  in  transverse  view  and  each  bears  a small  median  tubercle  on  its  anterior 
edge  and  is  separated  from  its  articulating  half-ring  by  a transversely  directed  furrow  which  bears  deep  apodemal 
pits  laterally.  Pleurae  parallel-sided  proximally,  tapering  slightly  over  the  distal  25%  where  they  are  deflected 
gently  downwards  and  rearwards.  Pleural  furrows  shallow,  each  directed  transversely  and  broadening  (exsag.) 
from  near  the  anteromesial  corner  of  the  pleura  such  that  the  posterior  band  tapers  abaxially  and  the  anterior 
band  expands  a little. 

Pygidium  broadly  similar  to  that  of  T.  ceriodes  angelini.  Rachis  composed  of  six,  possibly  seven  rings  and  the 
pleural  lobes  bear  up  to  three  poorly  defined  ribs. 


OWEN:  TRILOBITE  TRETASPIS 


725 


Discussion.  Hisinger  (1840)  described  two  species  of  Tretaspis,  ‘ Asaphus ’ seticornis  and  ‘A’.  cyllarus, 
from  the  Fjacka  Shale.  His  illustrations  of  both  show  the  development  of  four  complete  arcs  of  pits 
and  there  is  a well-developed  list  between  the  inner  and  the  outer  pairs  of  arcs  on  his  drawings  of  T. 
seticornis.  Dr.  J.  K.  Ingham  informs  me  (pers.  comm.  1976)  that  in  the  probable  syntypes  of  both 
species  and  all  other  available  specimens  from  the  Fjacka  Shale  at  Furudal  which  have  the  fringe 
preserved,  arcs  Ex,  Ix,  I2,  and  In  are  complete  and  a short  E2  is  developed  posteriorly.  Thus  it  seems 
reasonable  to  assume  that  this  is  indeed  the  case  with  the  syntypes  and,  in  order  to  stabilize  the 
species,  it  is  advocated  that  this  be  assumed  to  be  the  case.  Dr.  Ingham  has  examined  material  from 
Dalarna  described  by  Angelin  (1854)  as  T.  seticornis  and  considers  that  this  identification  probably  is 
correct.  Angelin’s  originals  of  T.  a ffinis  have  a complete  I3  arc  developed  and  thus  are  excluded  from 
T.  seticornis. 

Stormer  (1930)  assigned  a large  number  of  specimens  to  T.  seticornis  from  the  Fjacka  Shale  and 
various  horizons  in  Norway.  Many  of  these  are  reassigned  herein  to  T.  anderssoni  Stormer  and  T. 
hisingeri  sp.  nov.  It  is  clear  that  at  least  three  forms  are  present  in  the  Fjacka  Shale  and  so  references  to 
T.  seticornis  in  this  unit  by  Linnarsson  (1869)  and  Tornquist  (1884)  are  only  tentatively  included  in 
the  above  synonymy.  Further  discussion  of  material  previously  assigned  to  T.  seticornis  is  given 
below  in  the  discussions  of  T.  anderssoni  and  T.  hadelandica. 


Tretaspis  anderssoni  Stormer,  1945 
Plate  90,  figs.  5-10;  text-fig.  3 

?«oh1894  Trinucleus  seticornis  ( Hisinger);  Andersson,  p.  532,  figs.  1-5. 

1930  Tretaspis  seticornis  (Hisinger);  Stormer  (pars),  pi.  1 1,  figs.  2,  5;  pi.  12,  figs.  1-5;  pi.  13,  figs.  1,2, 
5-7;  ?pl.  14,  figs.  4,  5;  text-figs.  33 b,  c (pars),  d,  11c. 

71936  Tretaspis  seticornis  (Hisinger);  Asklund  (pars),  p.  4,  pi.  1,  fig.  4. 

1945  Tretaspis  seticornis  (Hisinger)  var.  anderssoni  Stormer,  p.  401,  pi.  1,  fig.  2. 

1959  Tretaspis  seticornis  (Hisinger);  Harrington  in  Moore,  text-figs.  52,  67. 

«o«1965  T.  seticornis  anderssoni  Stormer;  Cave,  p.  296  [?  = T.  hadelandica  brachystichus  Ingham]. 

1975  Tretaspis  seticornis  anderssoni  Stormer;  Hughes  et  al.,  p.  563,  pi.  4,  figs.  52,  53. 

1976  Tretaspis  seticornis  (Hisinger);  Miller,  text-fig.  2 h. 

1979  Tretaspis  seticornis  anderssoni  Stormer;  Owen  p.  253  text-fig.  8. 

71979  [specimens  resembling]  T.  hadelandica  Stormer;  Owen,  p.  253. 

Holotype.  A cranidium  (PM065196)  from  the  Frognoya  Shale,  on  Frognoya,  Ringerike. 

Material,  localities,  and  horizons.  Specimens  from  low  in  the  Frognoya  Shale  tentatively  compared  with  T. 
hadelandica  by  Owen  (1979)  probably  belong  in  T.  anderssoni  in  which  case  cephala,  cranidia,  lower  lamellae, 
and  pygidia  are  known  from  throughout  the  type  unit  on  Frognoya  and  from  the  overlying  Sorbakken 
Limestone  (except  the  lowest  9 m and  the  uppermost  17  m)  on  Frognoya  and  at  Norderhov,  Ringerike.  Two 
poorly  preserved  cranidia  from  the  Venstop  Shale  in  Skien-Langesund  may  belong  here  also. 

Description.  Cephalic  proportions  similar  to  those  of  T.  ceriodes  angelini.  The  fine  structure  of  the  median 
glabellar  tubercle  in  T.  anderssoni  was  described  by  Stormer  (1930,  p.  87,  text-fig.  37c;  pi.  11,  fig.  5;  pi.  13, 
figs.  5-7)  who  noted  that  it  bears  four  small  pits  arranged  as  at  the  corners  of  a square  and  a slightly  larger 
central  pit  which  may  bear  a fine  canal  opening.  Stormer  (1930,  pi.  12,  fig.  3;  pi.  13,  figs.  1,  2)  also  illustrated 
a lenticular  body  within  the  exoskeleton  of  the  lateral  tubercles  of  this  species.  On  the  external  surface  of  the 
glabella  and  genal  lobes  there  is  a weakly  developed  fine  reticulation  which  is  seen  faintly  on  a few  internal 
moulds. 

Fringe  narrow,  very  steeply  declined  except  laterally  where  a narrow  brim  is  developed.  A gentle  anterior  arch 
is  present.  The  details  of  fringe  pitting  are  given  on  text-fig.  3.  Two  distinct  sets  of  radii  are  present,  arcs  Ex,  Ix,  I2, 
and  In  are  complete  and  in  all  specimens  a short  E2  arc  is  developed  posteriorly  and  I3  is  developed 
anterolaterally  but  never  complete  mesially.  Arcs  Ix-E)  2 share  sulci  which  extend  to  between  bR5  and  bR14. 
The  limited  evidence  available  suggests  that  there  is  no  significant  difference  in  pit  development  between  early 
and  late  populations  of  T.  anderssoni. 

Hypostoma  and  thorax  unknown. 


726 


PALAEONTOLOGY,  VOLUME  23 


Pygidium  similar  to  that  of  T.  ceriodes  angelini.  Six  rachial  ring  furrows,  each  with  deep  apodemal  pits 
laterally,  are  seen  on  the  external  surface  of  the  rachis.  On  internal  moulds,  a seventh  pair  of  apodemal  pits  lies 
directly  in  front  of  the  pygidial  border.  Pleural  lobes  bear  four  weakly  developed  pairs  of  ribs,  the  posterior  two 
barely  discernible. 


SLI 

FS 


n=8 

n=8 


SLu 

SLI 

FS 


1=18 

i=5 


n=ll 


SLuNZMn=20 
SLI  MZMn=l4 
FS  MZ^n=20 


45- 

40- 

35. 


15  16  17  18  19  20 

Pits  in  In 


15. 

10. 

5. 


Pits  along  Pits  in  I3  Pits  Missing 

Post  Margin  Anteriorly  from  I3 


text-fig.  3.  Histograms  showing  the  range  of  variation  in  fringe  features  of  all 
available  specimens  of  Tretaspis  anderssoni  with  a comparison  of  samples  from 
the  Frognoya  Shale  (FS),  from  9-28  m above  the  base  of  the  overlying 
Sorbakken  Limestone  (SLI)  and  from  17  m below  the  top  of  this  unit  (SLu). 


EXPLANATION  OF  PLATE  90 

Figs.  1-4.  Tretaspis  seticornis  (Hisinger).  1,  PMO 103953,  dorsal  view  of  internal  mould  of  almost  complete 
specimen,  4-65  m above  base  of  Lower  Tretaspis  Shale,  Astaddammen,  Asker,  x 2\.  2,  PMO101553,  ventral 
view  of  part  of  cranidium,  Hogberg  Member  of  the  Solvang  Formation,  Frognoya,  Ringerike,  x4.  3, 
PMO  103954,  dorsal  view  of  cast  of  almost  complete  specimen,  Lower  Tretaspis  Shale,  Ole  Deviks  Vei,  Oslo, 
x 3.  4,  PMO103955,  anterolateral  view  of  cast  of  cranidium  showing  I3  developed  laterally,  1-65  m above 
base  of  Lower  Tretaspis  Shale,  same  locality  as  1,  x 4J. 

Figs.  5-10.  Tretaspis  anderssoni  Stormer.  5,PMO103956,  dorsal  view  of  internal  mould  of  pygidium,  17mbelow 
top  of  Sorbakken  Limestone,  Frognoya,  Ringerike,  x 6,8,9,  holotype,  PM065 1 96,  dorsal,  anterior,  and 
lateral  views  of  internal  mould  of  cephalon,  Frognoya  Shale,  Frognoya,  Ringerike,  x 3;  also  figured  by 
Stormer  (1945,  pi.  1,  fig.  2)  and  Hughes  et  al.  (1975,  pi.  4,  figs.  52,  53).  7,  PMO  H103,  posterolateral  view  of 
cephalon,  same  horizon  and  locality  as  6,  x 2 \\  also  figured  by  Stormer  (1930,  pi.  11,  fig.  5).  10,  PM080670, 
frontal  view  of  cast  of  cranidium,  Venstop  Shale,  Friefjord,  Skien-Langesund,  x 1\. 

Figs.  11-14.  Tretaspis  hisingeri  sp.  nov.  11,  PMO  H71,  frontal  view  of  partially  exfoliated  cephalon,  30-4-5  m 
below  top  of  Frognoya  Shale,  same  locality  as  6,  x 3£.  12-14,  PMO  H75,  dorsal,  lateral,  and  frontal  views  of 
partially  exfoliated  cranidium,  Frognoya  Shale,  same  locality  as  6,  x 3,  x 3|,  x 3£;  also  figured  by  Stormer 
(1930,  pi.  11,  fig.  3;  1945,  text-fig.  4). 


PLATE  90 


OWEN,  trilobite  Tretaspis 


728 


PALAEONTOLOGY,  VOLUME  23 


Discussion.  T.  anderssoni  differs  from  its  probable  ancestor,  T.  seticornis  in  having  a short  I3 
developed  in  all  specimens.  A broadly  similar  fringe  development  is  seen  in  a number  of  described 
taxa  and  their  interrelationships  are  discussed  below  under  T.  hadelandica. 

Stormer  (1945,  p.  40 1 ) considered  that  specimens  figured  by  Andersson  (1894)  as  T.  seticornis  from 
the  Lower  Johnstorp  Formation  (Pusgillian-?Cautleyan)  of  Hulderstad,  Oland,  Sweden,  probably 
belong  to  T.  anderssoni.  Examination  of  these  specimens  reveals  that  they  have  pit  counts  at  the  upper 
end  of,  or  even  beyond,  the  range  of  variation  seen  in  T.  anderssoni  from  Norway.  The  counts  in  these 
Riksmuseum,  Stockholm,  specimens  Ar21551  and  Ar21553  respectively  are  as  follows:  Ej  22,  20;  E2 
9 (710),  78;  In  c.  \1\,  18;  I3  5,  5.  Without  further  specimens  from  Oland  the  affinities  of  this  material 
must  remain  in  doubt.  Similarly,  a specimen  figured  by  Asklund  (1936)  from  the  Tretaspis  Beds  in 
Jemtland  has  a short  I3  but  its  affinities  must  await  the  description  of  further  specimens. 


Tretaspis  hisingeri  sp.  nov. 

Plate  90,  figs.  11-14;  Plate  91,  figs.  1-4;  text-fig.  4 

71887  Trinucleus  seticornis  (Hisinger);  Brogger,  p.  24. 

1 930  Tretaspis  seticornis  (Hisinger);  Stormer  {pars),  pi.  11,  figs.  1 , 3, 6, 7;  text-figs.  33c  (pars).  Til  a,  b,  40, 
47. 

1934  Tretaspis  seticornis ; Stormer  (pars),  p.  330. 

1945  Tretaspis  seticornis  (Hisinger)  forma  typica;  Stormer,  p.  401,  text-fig.  4. 

1970  T.  sp.  [?nov.];  Ingham,  p.  41. 

1975  T.  sp.  ?nov.;  Hughes  et  ai,  p.  563. 

1979  Tretaspis  sp.  nov.;  Owen,  p.  253,  text-fig.  8. 

Holotype.  An  almost  complete  specimen  (PMO  H51)  from  3-5-4-0  m below  the  top  of  the  Frognoya  Shale  on 
Frognoya,  Ringerike. 

Material,  localities,  and  horizons.  The  species  has  a limited  stratigraphical  distribution  and  is  known  from  all  but 
the  lowest  part  of  the  Frognoya  Shale  on  Frognoya  and  at  Hole  and  Norderhov,  and  also  between  9 and  14  m 
above  the  base  of  the  overlying  Sorbakken  Limestone  on  Frognoya,  Ringerike.  The  species  is  also  known  from 
the  upper  part  of  the  Lower  Tretaspis  Shale  at  Ole  Deviks  Vei  and  on  Bygdoy  and  Lindoya  in  Oslo,  the  Tretaspis 
Limestone  at  Nesbru,  Asker,  and  the  Venstop  Shale  in  Skien-Langesund. 

Diagnosis.  Very  narrow  fringe  has  El5  Il5  and  In  complete,  E2  short  and  a short  I2  present  in  the  vast  majority  of 
specimens  but  rarely  continuous  anteriorly,  and  in  some  instances  asymmetrically  distributed  about  the  sagittal 
line.  Two  distinct  sets  of  radii  mesially  and  where  I2  is  developed  but  laterally  In  is  in  phase  with  L-E,^. 


1^,,. 

16  17  18 

Pits  in  t 

19  20 

t|  10- 

^T- 

1 

A 

5 6 7 
Pits  along 
Posterior  Margin 


text-fig.  4.  Histograms  showing  the  range  of  variation  in  all  available  specimens  of 
Tretaspis  hisingeri  sp.  nov.  In  the  case  of  I2,  only  specimens  which  are  symmetrical 
about  the  sagittal  line  or  which  have  only  one  side  of  the  fringe  visible  are  included. 
An  additional  five  specimens  are  asymmetrical,  and  inclusion  of  the  right  or  left 
counts  with  the  data  shown  here  does  not  change  the  mean  value  although  the  left 
counts  increase  the  standard  deviation  to  lj.  The  number  of  pits  missing  from  I2 
anteriorly  from  these  specimens  is  the  same  for  both  right  and  left  sides,  and  thus 
are  incorporated  in  the  histogram  of  this  feature. 


OWEN:  TRILOBITE  TRETASPIS 


729 


Description.  The  glabella  and  genal  lobes  of  T.  hisingeri  differ  from  those  of  T.  ceriodes  angelini  only  in  having  the 
median  node  situated  a little  further  forward,  the  lateral  eye  tubercles  a little  closer  to  the  glabella,  and  in  most  of 
the  larger  holaspids  lacking  any  reticulation  on  the  external  surface  of  the  exoskeleton.  A specimen  of  meraspis 
degree  4,  however,  has  a very  strong  reticulation  on  both  glabella  and  genae  (pi.  91,  fig.  4).  Similar  reduction  in 
the  extent  and  intensity  of  reticulation  with  growth  in  trinucleids  is  well  documented  (Cech  1975).  Genal  spines 
extending  well  beyond  the  pygidium.  The  fringe  is  very  narrow  with  only  El5  Ix,  and  In  complete.  A short  E2  is 
developed  posteriorly  and  nearly  all  specimens  have  a few  pits  in  I2  which  is  rarely  continuous  frontally  (one 
specimen  out  of  twelve).  In  some  of  the  specimens  where  the  development  of  I2  can  be  seen  on  both  sides  of  the 
glabella  there  are  up  to  two  pits  less  on  one  side  than  on  the  other.  In  an  extreme  case  the  arc  is  absent  on  the  left 
side  but  contains  two  pits  on  the  right  (PI.  90,  fig.  1 1).  The  range  of  variation  in  fringe  features  is  illustrated  on 
text-fig.  4.  Arcs  In,  I x , Ex , and  (where  present)  E2  are  arranged  in  a single  set  of  radii  laterally  but  I j and  E l are  out 
of  phase  with  In  mesially  and  with  the  inner  two  I arcs  where  I2  is  developed. 

Hypostoma  unknown. 

Thorax  of  holaspis  similar  to  that  of  T.  seticornis.  That  of  the  meraspis  degree  4 noted  above  has  a narrower 
rachis  which  occupies  25%  (cf.  30%)  of  the  segment  width. 

Holaspid  pygidium  known  only  from  the  holotype  in  which  it  is  incomplete.  Rachis  bears  at  least  6 pairs  of 
apodemal  pits.  Meraspis  degree  4 pygidium  sub-semicircular  in  outline  with  a rachis  of  approximately  5 rings  of 
which  only  the  anterior  2 are  distinct. 

Discussion.  The  short  I2  development  distinguishes  T.  hisingeri  from  all  other  named  species.  T. 
hisingeri  succeeds  T.  seticornis  without  overlap  and  probably  was  derived  from  it  by  neoteny.  A very 
similar  form  in  which  I2  is  incomplete  but  more  extensive  than  in  T.  hisingeri  occurs  in  the  Fjacka 
Shale  in  Sweden,  and  it  too  succeeds  T.  seticornis  (J.  K.  Ingham,  pers.  comm.  1976). 

Tretaspis  hadelandica  hadelandica  Stormer,  1945 
Plate  91,  figs.  5-14;  Plate  92,  figs.  1,  2;  text-fig.  5 

1923  Trinucleus  sp.;  Holtedahl  in  Holtedahl  and  Schetelig,  p.  22. 

1945  Tretaspis  seticornis',  Stormer,  p.  384. 

1945  Tretaspis  seticornis  var.  hadelandica  Stormer,  pp.  384,  388,  406-407,  pi.  1,  figs.  3,  4. 

1945  Tretaspis  ceriodes  (Angelin);  Stormer,  pp.  387,  404-405,  pi.  4,  fig.  16. 

1945  Tretaspis  kiaeri  Stormer;  Stormer,  pp.  387,  406,  pi.  1,  fig.  11. 

1970  Tretaspis  hadelandica  hadelandica  Stormer;  Ingham,  text-fig.  17. 

1973  Tretaspis  seticornis',  Lauritzen,  p.  29. 

1978  Tretaspis  hadelandica  hadelandica  Stormer;  Owen,  pp.  11,  13,  14,  17. 

Holotype.  An  incomplete  cranidium  (PM065187)  probably  from  the  Gagnum  Limestone  Formation  south  of 
Gagnum,  Hadeland. 

Material,  localities,  and  horizons.  A few  complete  specimens  and  a large  number  of  disarticulated  skeletal 
elements  occur  abundantly  in  the  Gagnum  Shale  (except  the  lowest  part  in  northern  Hadeland)  and  Lunner 
Kirke  members  of  the  Lunner  Formation,  the  shales  of  this  formation  around  Lunner,  and  in  the  Gagnum 
Limestone  and  Kjorrven  formations.  The  species  is  rare  in  the  Grina  Shale  Member  of  the  Lunner  Formation. 
Fragmentary  museum  material  from  Nittedal  (precise  horizon  unclear),  between  Oslo  and  Hadeland,  may 
belong  here  also. 

Description.  Proportions  of  glabella  and  genal  lobes  very  similar  to  those  of  T.  ceriodes  angelini.  Specimens  from 
the  Gagnum  Shale  have  a well-developed  reticulation  on  the  external  surface  and  commonly  on  the  internal 
mould,  but  most  specimens  from  other  units  have  only  a subdued  reticulation  or  are  smooth.  Fringe  steeply 
declined  with  a slight  brim  developed  laterally.  Genal  spines  long,  diverging  rearwards  very  slightly.  All 
specimens  have  two  distinct  sets  of  radii,  and  arcs  E1;  Il5 12,  and  In  are  complete.  Three  morphs  are  recognized  on 
the  development  of  E2, 13,  and  I4  (Table  1)  and  the  distribution  of  pits  in  each  arc  is  shown  on  text-fig.  5.  Arc  I4  is 
absent  from  morphs  A and  B which  respectively  have  I3  incomplete  and  complete  posteriorly.  These  morphs 
occur  in  all  samples,  whereas  morph  C,  which  has  a short  I4  developed,  is  known  only  from  a few  populations 
from  the  lower  part  of  the  Gagnum  Shale  Member,  the  upper  part  of  the  Lunner  Formation  around  Lunner,  and 
from  the  Gagnum  Limestone.  In  one  specimen  (PI.  92,  figs.  1 , 2)  I3  is  complete  on  the  right  side  of  the  cranidium 
but  not  on  the  left,  an  asymmetry  which  encompasses  both  morph  A and  morph  B.  Samples  are  not  large  enough 


730 


PALAEONTOLOGY,  VOLUME  23 


B n=32  0,1  comP|e,e  posteriorly 


hh-l 


text-fig.  5.  Histograms  showing  the  range  of  variation  in  fringe  features  of  all 
available  specimens  of  Tretaspis  hadelandica  hadelandica  with  a comparison  of  the 
range,  mean,  and  sample  standard  deviation  of  the  three  morphs  (A,  B,  and  C) 
present  in  the  subspecies. 


to  enable  detailed  unit  by  unit  comparison  of  the  variation  in  each  morph  but  no  obvious  stratigraphical  changes 
are  apparent. 

Hypostoma  unknown.  Thorax  like  that  of  T.  seticornis. 

Pygidium  sub-semicircular  in  outline.  Rachis  crossed  by  5-7  furrows  each  bearing  apodemal  pits  laterally.  On 
the  ventral  surface  of  the  pygidium  there  are  up  to  ten  pairs  of  apodemes,  the  posterior  three  of  which  are  situated 
on  the  steeply  declined  pygidial  border.  Pleural  lobes  bear  three  low  ribs. 

Discussion.  When  present,  morph  C occurs  with  morphs  A and  B which  are  always  found  together. 
Their  great  similarity  in  pit  distribution  in  arcs  Ex,  In,  and  along  the  posterior  margin  argues  strongly 
for  these  morphs  being  no  more  than  broad  phenotypes  from  the  same  gene  pool.  Their  relative 
abundance,  however,  may  be  ecologically  controlled . T able  2 gives  the  relative  percentages  of  morphs 
present  in  the  stratigraphical  units  in  which  they  occur  in  measurable  abundance. 

3x2  and  2x2  contingency  tests  were  carried  out  on  the  specimen  numbers  used  to  calculate  these 
percentages.  The  latter  test  was  used  where  morph  C was  absent  from  both  samples  under 
examination,  or  where  expected  frequencies  of  morph  C were  less  than  5;  Yates’s  Correction  was 
applied  in  both  instances.  These  tests  show  that  the  Gagnum  Shale  abundances  are  significantly 
different  from  all  but  those  of  the  Kjorrven  Formation  at  the  0T%  level.  The  Kjorrven  Formation 


OWEN:  TRILOBITE  TRETASPIS 


73: 


table  2.  Percentages  of  each  morph  present  in  collections  of  T.  hadelandica  hadelandica  from  stratigraphical 

units  in  Hadeland 


Morph 

Gagnum 
Shale  Member 

Lunner  Kirke 
Member 

Lunner  Formation  above 
Lunner  Kirke  Member 

Gagnum 

Limestone 

Kjorrven 

Formation 

A 

55 

17 

18 

13 

47 

B 

31 

83 

75 

77 

53 

C 

14 

0 

7 

10 

0 

Number  106 

of  specimens 

35 

44 

39 

19 

abundances  differ  from  those  of  the  Gagnum  Shale  near  the  50%  level  which  is  not  significant,  and 
from  those  of  the  other  three  units  at  the  5%  level  which  is  considered  significant.  No  significant 
differences  are  present  between  the  remaining  three  units  where,  in  fact,  there  is  a high  degree  of 
correlation.  The  similarity  between  the  Gagnum  Shale  and  Kjorrven  Formation  abundances  is  the 
product  of  high  proportions  of  morph  A in  these  units.  It  may  be  noteworthy  that  both  units  have  a 
much  higher  trilobite  diversity  (measured  by  the  total  number  of  known  taxa)  than  the  others,  but 
speculation  on  the  reasons  for  this  similarity  in  morph  composition  would  be  very  unreliable  in  view 
of  the  small  sample  size  from  the  Kjorrven  Formation. 

T.  hadelandica  brachystichus  Ingham,  1970,  was  based  on  samples  from  the  Rawtheyan  Stage 
(Ashgill  Zones  5 and  6)  in  the  Cautley  area  of  northern  England  which  have  I3  incomplete  anteriorly 
and  posteriorly.  Ingham  also  tentatively  included  fragments  from  the  mid-Cautleyan  Stage  (Zone  3) 
in  this  subspecies  and  suggested  that  specimens  from  the  Gagnum  Shale  assigned  to  T.  ceriodes  by 
Stormer  may  belong  to  the  north  of  England  form.  These  Gagnum  Shale  specimens  are  assigned  to  T. 
hadelandica  hadelandica  morph  A herein.  Ingham’s  material  and  specimens  assigned  to  T. 
hadelandica  brachystichus  by  Price  (1973,  1977)  and  Cocks  and  Price  (1975)  from  the  uppermost  part 
of  the  Sholeshook  Limestone  and  lower  part  of  the  Slade  and  Redhill  Mudstone  (mid-Ashgill)  in 
south  Wales,  have  a range  of  variation  which  overlaps  that  seen  in  the  Norwegian  morph  A (text-fig. 
6).  In  the  case  of  arcs  E2  and  Ej  and  the  number  of  pits  along  the  posterior  margin,  the  range  and,  in 
the  E arcs,  the  mean  is  higher  than  that  of  the  Norwegian  morph.  The  variation  in  number  of  pits  in 
I3,  however,  overlaps  at  the  lower  end  of  that  seen  in  T.  hadelandica  hadelandica  morph  A and  is 
closer  to  that  of  T.  anderssoni  which,  in  all  these  characters,  has  a range  of  variation  which  overlaps 
only  at  its  upper  end  with  that  of  T.  hadelandica  hadelandica  morph  A (text-fig.  6). 

T.  corner  gens  Dean,  1961 , was  described  originally  from  Pusgillian  strata  in  the  Cross  Fell  Inlier  in 
northern  England  and  subsequently  by  Ingham  (1970)  from  Pusgillian  and  lower  Cautleyan  (Ashgill 
Zone  1)  strata  at  Cautley  and  by  McNamara  (1979)  from  mid-Cautleyan  (Zone  2 and  lowest  part  of 
Zone  3)  strata  in  the  English  Lake  District.  As  noted  by  McNamara  ( 1 979,  p.  62),  the  limited  evidence 
available  suggests  that  there  is  a progressive  reduction  in  I pits  with  the  Cross  Fell  specimens  having  a 
short  (up  to  ten  pits)  I4  arc,  some  specimens  lacking  this  arc  in  the  Cautley  material,  and  in  all  the 
Lake  District  specimens  this  arc  is  not  developed.  This  trend  is  continued  in  Ashgill  Zone  3 in  the 
Lake  District  with  T.  convergens  deliquus  McNamara,  1979;  I3  becoming  incomplete  anteriorly  and 
then  laterally.  The  earliest  examples  with  T.  convergens  deliquus  morphology  occur  with  specimens 
with  I3  complete  anteriorly  (K.  J.  McNamara,  pers.  comm.  1979). 

As  text-fig.  6 shows,  the  over-all  pit  distribution  and  the  broad  range  of  variation  seen  in  the 
subspecies  of  T.  convergens  is  very  similar  to  that  of  T.  hadelandica  hadelandica  and  consequently  the 
English  forms  are  regarded  as  subspecies  of  T.  hadelandica.  It  seems  reasonable  to  suggest  that 
Ingham’s  indeterminate  specimens  from  Zone  3 are,  in  fact,  T.  hadelandica  deliquus  and  that  the 
progressive  decrease  in  pit  number  in  I3  documented  by  McNamara  continued,  giving  rise  to  T. 
hadelandica  brachystichus.  Moreover,  if  the  Zone  3 material  from  Cautley  is  indeed  T.  hadelandica 
deliquus,  the  restriction  of  T.  hadelandica  brachystichus  to  Zones  5 and  6 (i.e.  lower  Rawtheyan)  in 
northern  England  would  add  weight  to  Ingham’s  suggestion  ( 1 977,  p.  118)  that  the  uppermost  part  of 
the  Sholeshook  Limestone  is  early  Rawtheyan  in  age.  T.  convergens  has  been  recorded  from  lower 


732  PALAEONTOLOGY,  VOLUME  23 

Ashgill  strata  at  Girvan,  south-west  Scotland  (Ingham  1970,  p.  46),  but  the  affinities  of  this  material 
are  not  known. 

The  succession  of  subspecies  of  T.  hadelandica  in  northern  England  seems  to  represent  a single 
local  stock  and  the  above  revision  is  based  on  this.  An  alternative,  but  more  contrived  hypothesis 
would  be  the  ecological  replacement  of  T.  hadelandica  hadelandica  morphs.  Thus  the  Norwegian 
morph  C resembles  early  T.  hadelandica  convergens,  morph  B resembles  late  T.  hadelandica 
convergens,  and  early  T.  hadelandica  deliquus  and  morph  A resembles  T.  hadelandica  brachystichus. 

As  far  as  morph  A is  concerned,  the  absence  of  E2  in  some  specimens,  the  high  percentage  of 
individuals  in  which  I3  is  continuous  frontally,  and  the  fairly  limited  overlap  in  number  of  pits  in  I3 
serves  to  distinguish  it  from  T.  hadelandica  brachystichus.  Chi-squared  tests  show  that  the  ranges  in  I3 
pits  and  pits  missing  anteriorly  from  this  arc  are  distinct  at  the  0T%  level,  even  when  Ingham’s 
samples  from  Zones  5 and  6 are  considered  together.  There  is  only  a limited  amount  of  information 
on  T.  hadelandica  convergens  and  T.  hadelandica  hadelandica  morph  C from  the  Gagnum  Shale 
but  this  suggests  that  the  English  form  commonly  has  more  pits  in  E3  (19^-22^  cf.  18-1 9^)  and 
along  the  posterior  margin  (7-12  cf.  6-7),  but  fewer  in  E2  (6-11  cf.  10-13)  and  in  all  cases  I3  is 
complete  whereas  it  is  incomplete  posteriorly  in  33%  (of  nine  specimens)  of  the  Norwegian  morph  C. 
The  English  form  is  also  distinguished  by  its  more  swollen  pseudofrontal  lobe.  Although  later 


n % Present  n 

T.  hadelandica  brachystichus  S.  Woles  23  100  23 

T.  hadelandica  brachystichus  Zone  6 19  100  19 

T hadelandica  brachystichus  Zone  5 36  100  36 

T hadelandica  deliquus  1 1 100  1 1 

T hadelandica  convergens  7 100  7 

T.  hadelandica  hadelandica  Morph  B 30  100  35 

T hadelandica  hadelandica  totol  62  94  69 

T.  hadelandica  hadelandica  Morph  A 18  82  22 

T.  anderssoni  38  100  76 


%Complete 

0 

0 

0 

0 

0 

17 

7 

0 

0 


20  21  22  23 


T.  had.  bra.  S.  Wales 
T.  had.  bra  Z 5 a 6 
T.  had.  del. 

T.  had.  conv. 

T.  had.  had.  B 
T.  had.  had.  total 
T.  had.  had.  A 
T anderssoni  \ 


VZ  I 2 3 

Pits  Missing  Anteriorly  from  I3 
( where  incomplete  here ) % complete 

post 

T.  had.  bra.  S.  Wales  0 

T.  had.  bra  Z.  6 0 

T.  had.  bra  Z . 5 0 

T.  had.  del.  1 1 

T.  had.  conv.  100 


T.  had.  had.  B 
T had.  had  total 
T.  had.  had.  A 
T.t 


text-fig.  6.  Range,  mean,  and  sample  standard  deviation  of  selected  fringe  characters  of  members  of 
the  Tretaspis  seticornis  group  in  which  I3  is  incomplete  in  at  least  some  individuals.  Data  for  T. 
hadelandica  convergens,  T.  h.  deliquus,  and  T.  h.  brachystichus  based  on  histograms  given  by  Ingham 
(1970),  McNamara  (1979),  and  Price  (1977). 


OWEN:  TRILOBITE  TRETASPIS 


733 


populations  of  T.  hadelandica  convergens  and  early  T.  hadelandica  deliquus  lack  I4,  they  differ  from  T. 
hadelandica  hadelandica  morph  B in  always  having  I3  complete  frontally.  It  seems  most  likely, 
therefore,  that  T.  hadelandica  hadelandica  with  its  broad  range  of  variation  (morphs  A,  B,  and  C)  and 
the  British  series  of  subspecies  with,  at  any  one  level,  a much  narrower  range  of  variation  were  at  most 
connected  by  a series  of  clines  throughout  much  of  the  Ashgill. 

T.  clarkei  Cooper  (in  Schuchert  and  Cooper,  1930)  from  Ashgill  units  in  Quebec,  Canada,  has  two 
distinct  sets  of  radii  and  thus  belongs  to  the  T.  seticornis  group  and  is  not  a synonym  of  T.  ceriodes  (cf. 
Whittington  1941,  p.  29;  Lesperance  1968,  p.  813;  Bolton  1970,  pp.  35-36).  The  holotype  from 
the  Whitehead  Formation  and  specimens  figured  by  Bolton  (1970,  pi.  6,  figs.  12,  15,  17,  19)  from 
the  Vaureal  Formation  have  I3  incomplete  posteriorly  at  least.  Of  the  three  specimens  from  the 
Whitehead  Formation  in  the  Hunterian  Museum,  two  (HM  A4319;  4320)  have  eight  pits  in  I3  which 
is  incomplete  anteriorly.  A third  specimen  (HM  A4321)  has  twelve  pits  in  I3  which  is  complete 
anteriorly  and  three  pits  in  I4.  It  is  not  known  whether  the  specimens  are  from  the  same  horizon  but 
all  fringe  features  fall  within  the  range  seen  in  T.  hadelandica.  Detailed  sampling  of  T.  clarkei 
populations  is  needed  before  its  affinities  can  be  fully  determined. 

As  is  noted  in  the  discussion  of  T.  seticornis,  specimens  of  Tretaspis  with  arc  I3  developed  are 
known  from  the  Fjacka  Shale  in  Sweden.  In  addition  to  Angelin’s  material  of  T.  affinis , which  has 
this  arc  complete,  other  specimens  in  the  Riksmuseum,  Stockholm,  have  I3  incomplete  but,  in  some 
cases,  extensive  (J.  K.  Ingham,  pers.  comm.  1976).  Dr.  Ingham  has  also  examined  a specimen  from  the 
Slandrom  Limestone  (probably  early  Pusgillian)  in  the  Siljan  district  (Jaanusson  and  Martna  1948, 
p.  187)  which  has  a short  I3  and  a coarsely  reticulate  glabella  and  genal  lobes.  Dr.  P.  J.  Brenchley  of 
Liverpool  University  has  sent  me  a specimen  resembling  T.  hadelandica  hadelandica  morph  B from 
the  flank  facies  of  the  Boda  Limestone  (Ashgill)  in  the  Siljan  district  and  this  is  the  only  specimen  of 
Tretaspis  known  from  these  beds,  and  the  genus  is  not  known  from  the  Boda  Limestone  itself.  The 
Swedish  species  of  Tretaspis  are  being  revised  by  Dr.  Ingham  who  has  taken  well-localized  samples 
from  the  Fjacka  Shale. 

There  is  a great  deal  of  other  material  of  the  T.  seticornis  group  with  an  incomplete  I3  and  in  need  of 
modern  study.  This  includes  specimens  from  Ashgill  units  in  Poland  ascribed  to  T.  seticornis  by 
Kielan  (1957,  1960)  and  Tomczyk  (1962),  and  material  from  the  Kraluv  Dvur  Formation  (mid- 
Ashgill)  in  Bohemia  examined  by  the  writer  in  the  collections  of  the  British  Museum  (Natural 
History).  Specimens  from  this  latter  unit  were  referred  to  T.  seticornis  by  Havlicek  and  Vanek  (1966) 
and  Pribyl  and  Vanek  (1969).  Ingham  (1970,  pp.  41, 49)  noted  that  specimens  which  Lamont  (1935, 
1941)  assigned  to  T.  seticornis  from  the  Lower  Drummuck  Group  (Cautleyan)  at  Girvan  has  a short 
I3  and  was  termed  T.  sp.  by  Hughes  etal.  (1975,  p.  563).  Price  (1977,  p.  786)  noted  a similarity  between 
an  unnamed  form  from  low  in  the  Slade  and  Redhill  Mudstones  and  this  species.  Dr.  Ingham  informs 
me  (pers.  comm.  1976)  that  T.  seticornis  of  Portlock  (1843)  and  Fearnsides,  Elies,  and  Smith  (1907) 
from  the  Killey  Bridge  Beds  (low  Cautleyan)  in  Pomeroy,  Ireland,  may  well  prove  synonymous  with 
the  broadly  coeval  Lower  Drummuck  Group  form  as  both  have  a short  I3,  a very  extensive  E2,  and 
large  lateral  eye  tubercles  quite  close  to  the  glabella.  Moreover,  T.  sp.  probably  gave  rise  to  T. 
persulcatus  (Reed,  1935)  from  the  Upper  Drummuck  Group  (late  Rawtheyan)  in  which  E2  is 
complete  and  the  girder  is  indistinct  posteriorly  where  an  external  pseudogirder  is  developed  between 
Ex  and  E2  (see  Ingham,  1970,  p.  44). 

Schmidt  (1894)  assigned  specimens  to  T.  seticornis  from  the  Lykholm  Group  (late  Caradoc  to 
Ashgill)  in  Estonia,  and  Jaanusson  (1956,  pp.  379,  383)  listed  the  species  from  the  lower  part  of  the 
group,  the  Nabala  Formation  (late  Caradoc).  It  is  not  known  whether  the  material  referred  to  by 
Jaanusson  is  from  the  same  beds  as  Schmidt’s  specimens,  one  of  which  (1894,  pi.  5,  fig.  22)  is 
illustrated  as  having  I3  complete  posteriorly,  but  it  is  not  clear  whether  two  sets  of  radii  are 
developed.  Assuming  that  existing  correlations  are  correct,  the  Estonian  specimens  listed  by 
Jaanusson  would  prove  the  oldest  record  of  the  T.  seticornis  group  should  they  prove  correctly 
ascribed  to  it. 


734 


PALAEONTOLOGY,  VOLUME  23 


Tretaspis  latilimbus  (Linnarsson,  1869)  norvegicus  subsp.  nov. 

Plate  92,  figs.  3-7;  text-fig.  7 

1887  Trinucleus  seticornis  (Hisinger)  (?)  var.;  Brogger,  p.  26. 

1887  Trinucleus  conf.  seticornis ; Brogger,  p.  29. 

1887  Trinucleus-,  Brogger,  p.  30. 

1887  Trinucleus  Wahlenbergi;  Brogger,  p.  31. 

71887  Trinucleus  Wahlenbergi  Rouault;  Brogger,  p.  32. 

1897  Trinucleus  Wahlenbergi  Rouault;  Kiaer,  p.  33  [Upper  Isotelus  Limestone,  ?‘5a’]. 

1930  Tretaspis  latilimbus  (Linnarsson);  Stormer  (pars),  pp.  67-69  [Tretaspis  Limestone  specimens  only], 
pi.  11,  figs.  8,  ?9,  10,  11;  text-figs.  33/,  ?g,  non  e [=  T.  anderssoni ],  non  34 d [=  T.  latilimbus 
latilimbus \. 

1934  Tretaspis  latilimbus-,  Stormer,  p.  330. 

1945  Tretaspis  latilimbus  (Linnarsson);  Stormer,  p.  403,  pi.  1,  fig.  9. 

Holotype.  A cephalon  (PMOl  1751)  from  the  Tretaspis  Limestone  on  Lindoya,  Oslo. 

Material,  localities,  and  horizons.  A great  deal  of  very  fragmentary  material  and  rarer  more  complete  specimens 
occur  at  various  levels  in  Oslo-Asker:  Tretaspis  Limestone  on  Langara,  Lindoya,  Ostoya,  and  Treneholmen; 
Upper  Tretaspis  Shale  on  Hovedoya  and  Nakholmen;  Upper  Isotelus  Limestone  on  Hovedoya,  Langoyene, 
Lindoya,  and  Skjaerholmen;  all  but  the  upper  few  metres  of  the  Husbergoya  Shale  Formation  on  Hovedoya  and 
possibly  Husbergoya,  Rambergoya,  and  Langoyene.  A specimen  in  limestone  (?Heroya  Limestone)  from  the 
Skien-Langesund  district  probably  belongs  here  also. 

Diagnosis.  Arcs  El5  I13,  and  In  complete.  I4  short  to  complete,  E2  present  in  41%  of  thirty-four  specimens. 
Reticulation  on  external  surface  of  glabella  and  genae  subdued. 

Description.  Proportions  of  glabella  and  genal  lobes  similar  to  those  of  T.  ceriodes  angelini.  There  is  a fine, 
subdued  reticulation  on  the  external  surface  of  the  mesial  part  of  the  glabella  and  the  adaxial  parts  of  the  genal 
lobes,  but  they  are  smooth  on  internal  moulds.  Arcs  Ex,  I^,  and  In  are  complete,  and  I4  is  developed  in  all 
specimens,  most  having  3-1 1^  pits  (twenty-four  specimens)  in  this  arc  but  one  extreme  specimen  from  the 
Tretaspis  Limestone  has  this  arc  complete.  Two  morphs  (A  and  B)  are  defined  on  the  absence  or  presence 


EXPLANATION  OF  PLATE  91 

Figs.  1-4.  Tretaspis  hisingeri  sp.  nov.  1,  2,  holotype,  PMO  H51,  dorsal  and  lateral  views  of  partially  exfoliated 
almost  complete  specimen,  3-5-4-0  m below  top  of  Frognoya  Shale,  Frognoya,  Ringerike,  x2£,  x2;  also 
figured  by  Stormer  (1930,  text-fig.  47).  2,  PMO80613,  frontal  view  of  partially  exfoliated  cranidium, 
Frognoya  Shale,  Ringsasen,  Norderhov,  Ringerike,  x2.  4,  PMO  103957,  dorsal  view  of  cast  of  complete 
meraspis  degree  4,  7-91-7-94  m above  base  of  Lower  Tretaspis  Shale,  Ole  Deviks  Vei,  Oslo,  x 12^. 

Figs.  5-14.  Tretaspis  hadelandica  hadelandica  Stormer.  5,  8,  11,  holotype,  morph  B,  PM065187,  dorsal,  frontal, 
and  lateral  views  of  partially  exfoliated  cephalon,  probably  from  the  Gagnum  Limestone  Formation,  south  of 
Gagnum,  Hadeland,  x 2;  also  figured  by  Stormer  (1945,  pi.  1,  fig.  4).  6,  morph  B,  PM098489,  dorsal  view  of 
lower  lamella  external  to  girder  showing  E2  complete  frontally,  upper  part  of  Lunner  Formation,  Kjevlingen, 
Hadeland,  x 3-2.  7,  morph  C,  PMO  103958,  anterolateral  view  of  partially  exfoliated  cranidium,  Gagnum 
Limestone  Formation,  500  m south-east  of  Lunner  Bakken,  Hadeland,  x 4 \.  9,  morph  A,  PM099537, 
oblique  anterolateral  view  of  cast  of  cranidium,  7- 1 -7-2  m below  top  of  Gagnum  Shale  Member  of  the  Lunner 
Formation,  75  m south  of  Roko,  Hadeland,  x 5.  10,  PMO103959,  dorsal  view  of  unwhitened  pygidium, 
lower  part  of  Lunner  Formation,  400  m east-south-east  of  Lunner  Kirke,  Hadeland,  x 8.  12,  PMO101483, 
dorsal  view  of  internal  mould  of  pygidium,  Gagnum  Limestone  Formation,  Ballangrud,  Hadeland,  x 4.  13, 
PMO  103960,  dorsal  view  of  unwhitened  thorax  and  pygidium,  lower  part  of  Lunner  Formation,  Haga, 
Hadeland,  x4.  14,  morph  A,  PM065193,  dorsal  view  of  partially  exfoliated  almost  complete  specimen, 
Gagnum  Shale  Member  of  the  Lunner  Formation,  Gagnum,  Hadeland,  x 3£;  also  figured  by  Stormer  (1945, 
pi.  4,  fig.  16). 


PLATE  91 


OWEN,  trilobite  Tretaspis 


736 


PALAEONTOLOGY,  VOLUME  23 


respectively  of  E2  which  occurs  in  41%  of  the  thirty-four  specimens  in  which  this  feature  could  be  determined 
(Table  1).  As  in  all  species  of  Tretaspis  the  most  posterior  one  or  two  E,  pits  always  lack  equivalent  E2  pits.  The 
range  of  variation  in  fringe  pitting  is  illustrated  on  text-fig.  7.  Arcs  Ix,  El5  and,  where  present,  E2  are  out  of  phase 
with  the  remaining  I arcs  and  share  sulci  which  extend  to  the  anterolateral  part  of  the  fringe  or  even  to  the  zone  of 
complication.  The  available  samples  are  too  small  to  detect  differences  from  rock  unit  to  rock  unit  and  both 
morphs  are  known  from  all  but  the  lower  part  of  the  Husbergoya  Formation.  Personally  collected  material  from 
the  Upper  Tretaspis  Shale  shows  both  morphs  in  the  same  bed. 

Hypostoma  unknown.  Thorax  similar  to  that  of  T.  seticornis. 

Pygidium  only  known  with  certainty  from  a few  fragments.  The  one  figured  by  Stormer  (1930,  pi.  11,  fig.  9) 
may  belong  here  or  to  T.  sortita  broeggeri  (see  below)  as  the  precise  horizon  in  the  Husbergoya  Formation  is  not 
known.  This  specimen  has  ten  pairs  of  apodemal  pits,  the  posterior  three  lying  on  the  pygidial  border. 


3 5 7 

Pits  in  E2 
( morph  B only ) 


JTfl- 


n R ! i i n //i— r 


Pits  Missing  Pits  along 

Anteriorly  from  I4  Posterior  Margin 


text-fig.  7.  Histograms  showing  range  of  variation  in  fringe  features  of  all 
available  specimens  of  Tretaspis  latilimbus  norvegicus  with  a comparison,  where 
possible,  of  the  range,  mean,  and  sample  standard  deviation  of  the  two  morphs 
(A  and  B)  present  in  the  subspecies.  In  many  instances  the  samples  are  too  small 
for  reliable  standard  deviations  or  even  means  to  be  calculated. 


Discussion.  Ingham  (1970,  p.  50,  text-fig.  18a,  b)  chose  a lectotype  from  Linnarsson’s  original 
material  of  T.  latilimbus  from  the  Upper  Johnstorp  Formation  (Rawtheyan)  of  Vastergotland, 
Sweden,  and  he  figured  a number  of  topotypes  (1970,  text-fig.  18c-/).  Dr.  Ingham  has  allowed  me  to 
collate  some  of  his  data  on  topotype  material  of  the  Swedish  form  in  the  collections  of  the 
Riksmuseum,  Stockholm.  All  these  specimens  have  an  incomplete  I4  arc  (2-11  pits  in  forty-two 
specimens)  which  is  not  continuous  mesially.  Where  the  E arc  development  is  sufficiently  well 
preserved,  only  one  specimen  out  of  thirty-five  is  seen  to  have  pits  in  E2  and  thus  the  vast  majority 
correspond  to  the  development  seen  in  T.  latilimbus  norvegicus  morph  A.  The  range  of  variation  seen 
in  the  pit  distribution  of  other  arcs  is  similar  to  that  of  the  Norwegian  material  and  it  is  most  likely 
that  the  Swedish  form  is  simply  a geographical  subspecies  of  T.  latilimbus  norvegicus  in  which  morph 
B has  been  virtually  excluded.  Fragments  of  Tretaspis  from  the  Ulunda  Formation  (Rawtheyan)  in 
Vastergotland  have  a pit  development  similar  to  that  of  T.  latilimbus  norvegicus  morph  B (J.  K. 
Ingham,  pers.  comm.  1976). 

T.  Tatilimbus'  distichus  Ingham  (1970,  p.  50,  pi.  7,  figs.  8-16,  text-figs.  14 g,  16)  was  based  on 
material  from  the  Rawtheyan  Stage  (Ashgill  Zone  7)  in  the  Cautley  district  of  northern  England  and 
is  characterized  by  the  presence  of  a short  I4  and  seven  to  ten  pits  in  E2.  It  thus  resembles  T.  latilimbus 
norvegicus  morph  B and  might  be  regarded  as  being  a subspecies  which  developed  in  the  same  way  as 
T.  latilimbus  latilimbus.  However,  Ingham  (1970,  p.  50)  suggested  that  T.  ‘ latilimbus ’ distichus  may 


OWEN:  TRILOBITE  TRETASPIS 


Til 


have  been  derived  from  T.  hadelandica  brachystichus  with  the  completion  of  I3  and  the  development 
of  a short  I4.  Indeed,  one  specimen  of  the  latter  was  noted  by  Ingham  to  have  a pit  in  I4.  When  the 
ranges  in  variation  in  E2  and  I3  in  T.  hadelandica  brachystichus  are  considered  for  Zones  5 and  6 
separately  (text-fig.  6;  Ingham  1970,  text-fig.  16)  there  is  a suggestion  of  a trend  towards  the  condition 
seen  in  T.  ' latilimbus’  distichus.  Moreover,  McNamara  (1979,  p.  63)  has  noted  the  occurrence  of 
specimens  which  he  terms  T.  aff.  latilimbus  distichus  from  the  White  Limestone  (top  of  Zone  6)  in  the 
Lake  District  which  he  considers  to  be  intermediate  between  T.  hadelandica  brachystichus  and  T. 
‘‘latilimbus’  distichus.  Dr.  McNamara  informs  me  (pers.  comm.  1979)  that  the  White  Limestone  form 
has  I3  complete  posteriorly  and  in  some  specimens  there  is  a single  pit  in  I4.  Thus  it  seems  likely  that 
the  Zone  7 form  is  not  directly  related  to  T.  latilimbus  and  ultimately  may  best  be  considered  a 
stratigraphical  subspecies  of  T.  hadelandica. 

The  origins  of  T.  latilimbus  are  not  clear  but  T.  hadelandica  hadelandica  seems  to  be  the  most  likely 
ancestor. 


Tretaspis  sortita  (Reed,  1935)  broeggeri  Stormer,  1945 
Plate  92,  figs.  8-11,  13,  14;  text-fig.  8 
71887  Trinucleus  Wahlenbergi  Rouault;  Brogger,  p.  32. 

71897  Trinucleus  Wahlenbergi  Rouault;  Kiaer,  pp.  32  {pars,  ‘4d§’  specimens  only),  73. 

1945  Tretaspis  latilimba  (Linnarsson)  var.  broggeri  Stormer,  p.  403,  pi.  1,  fig.  10. 

1979  Tretaspis  sortita  (Reed)  broeggeri  Stormer;  Owen,  p.  257. 

Holotype.  An  incomplete  cephalon  (PMO 11957)  from  the  upper  part  of  the  Husbergoya  Shale  Formation  on 
Skjaerholmen,  Oslo. 

Material,  localities,  and  horizons.  Cephala,  cranidia,  and  rare  thoracic  segments  are  known  from  the  upper  few 
metres  of  the  Husbergoya  Formation  in  Oslo,  on  the  islands  of  Skjaerholmen,  Husbergoya  (upper  2 m), 
Hovedoya  (upper  2-5  m),  South  Langoyene  (upper  5 m),  Lindoya,  and  Gressholmen. 

Description.  Glabella  and  genal  lobes  similar  to  those  of  T.  ceriodes  angelini  but  bear  a variably  developed, 
subdued,  fine  reticulation  on  the  external  surface  (on  the  genal  lobes  this  is  restricted  to  the  posterior  parts)  and 
are  smooth  on  internal  moulds.  Fringe  steeply  declined  with  a well-developed  anterior  arch  and  a distinct  brim 
laterally.  Arcs  E1;  I1_3,  and  In  are  complete,  I4  is  incomplete,  and  three  morphs  are  recognized  on  the  basis  of  the 
E2  and  1 5 development  (Table  1)  thus:  morph  A lacks  E2  and  I5,  morph  B lacks  I5  but  has  a short  E2,  morph  C 
has  both  E2  and  I5  present  but  incomplete.  The  range  of  variation  in  pit  development  is  shown  on  text-fig.  8. 
There  are  too  few  specimens  to  give  meaningful  comparisons  of  some  of  the  parameters  in  the  three  morphs 
separately,  and  in  some  instances  morphs  A and  B are  considered  together  on  text-fig.  8.  Arc  E2  is  irregularly 
developed  in  a few  specimens  (e.g.  PI.  92,  figs.  8,  10)  but  in  most  cases  where  it  is  present  it  is  restricted  to  the 
posterior  part  of  the  fringe.  In  all  morphs,  arcs  Ij-Ej  share  sulci  to  bR2-8  (mean  bR5,  sample  standard  deviation 
1,  eighteen  specimens)  and  distinct  lists  are  developed  between  most  arcs  over  the  whole  fringe  except  Ej-I,, 
where  they  share  sulci  and  between  Ej  and  E2  and  also  I5  and  In.  Two  sets  of  radii  are  developed. 

Hypostoma  not  known.  Thorax  and  pygidium  known  only  from  a few  poorly  preserved  fragments. 

Discussion.  The  holotype  of  T.  sortita  broeggeri  has  eleven  pits  in  E2  and  one  in  I5,  and  thus  is  of 
morph  C type.  Morphs  A and  B are  indistinguishable  from  the  two  morphs  constituting  T.  latilimbus 
norvegicus  although  their  relative  abundances  are  very  different  with  E2  being  developed  much  more 
commonly  in  T.  sortita  broeggeri.  Stormer’s  subspecies  probably  was  derived  from  T.  latilimbus 
norvegicus  with  the  development  of  1 5 in  some  individuals  and  replaces  the  earlier  form  quite  abruptly 
in  the  Husbergoya  Formation,  although  it  is  not  possible  to  assign  unequivocally  isolated  specimens 
of  morphs  A and  B to  either  form. 

T.  sortita  sortita  (Reed,  1935,  pp.  3-6,  pi.  1,  figs.  4-10;  see  also  Begg  1944,  pp.  114,  115,  pi.  5,  figs. 
2-7)  was  based  on  material  from  the  Upper  Drummuck  Group  (late  Rawtheyan)  at  Girvan,  south- 
west Scotland.  A complete  topotype  specimen  was  figured  by  Ingham  (1970,  pi.  8,  fig.  1)  who  noted 
(1970,  p.  50)  that  the  Scottish  form  has  an  incomplete  E2,  an  extensive  but  incomplete  I4,  and  a few 
pits  in  I5.  Dr.  Ingham  has  informed  me  (pers.  comm.  1976)  that  the  vast  majority  of  specimens  from 


738 


PALAEONTOLOGY,  VOLUME  23 


Pits  in  E2 


C 

A+B 

i°-| 

5- 


1/2  1 2 3 

Pits  Missing 
Anteriorly  from  I4 


Pits  in  I4 


nb  42%  of  31  specimens  of  morphs  AH-B 
and  93  % of  30  specimens  of  morph  C 
have  this  arc  complete  anteriorly 


Pits  in  In 


Pits  Missing 
Anteriorly  from  1 5 


Pits  along 
Posterior  Margin 


text-fig.  8.  Histograms  showing  the  range  of  variation  in  fringe  features  of 
all  available  specimens  of  Tretaspis  sort  it  a broeggeri  with  a comparison  of  the 
range,  mean,  and  sample  standard  deviation  of  the  three  morphs  (A,  B,  and 
C)  present  in  the  subspecies.  Owing  to  the  limited  amount  of  data  for  morphs 
A and  B,  these  are  considered  together  in  most  instances. 


EXPLANATION  OF  PLATE  92 

Figs.  1,2.  Tretaspis  hadelandica  hadelandica  Stormer.  PMO 103961,  right  and  left  lateral  views  of  internal  mould 
of  cranidium  showing  asymmetrical  I3  development,  lower  part  of  Gagnum  Shale  Member  of  the  Lunner 
Formation,  200  m north  of  Aslund,  Hadeland,  x 8. 

Figs.  3-7.  Tretaspis  latilimbus  (Linnarsson)  norvegicus  subsp.  nov.  3,  holotype,  ?morph  A,  PMOl  1751,  antero- 
lateral view  of  incomplete  cephalon,  Tretaspis  Limestone,  Lindoya,  Oslo,  x 4;  also  figured  by  Stormer  (1945, 
pi.  1 , fig.  9).  4,  morph  B,  PMO  10 1 551,  anterolateral  view  of  partially  exfoliated  cranidium,  same  horizon  as  3, 
west  Rambergoya,  Oslo,  x3.  5,  morph  A,  PMO80518,  oblique  anterolateral  view  of  internal  mould  of 
cephalon,  Husbergoya  Shale  Formation,  North  Langoyene,  Oslo,  x 3.  6,  morph  A,  PMO  1 03962,  dorsal  view 
of  internal  mould  of  cephalon  and  part  of  thorax,  Upper  Tretaspis  Shale,  north  Hovedoya,  Oslo.  7,  morph  B, 
PMO80573,  oblique  posterolateral  view  of  cephalon,  same  horizon  as  3,  Ostoya,  Baerum,  x 3. 

Figs.  8-11,  13,  14.  Tretaspis  sortita  (Reed)  broeggeri  Stormer.  8,  morph  B,  PMO31010,  anterolateral  view  of 
internal  mould  of  incomplete  cephalon  showing  irregular  E2  development,  upper  part  of  Husbergoya  Shale 
Formation,  South  Langoyene,  Oslo,  x 2\.  9,  holotype,  morph  C,  PMOl  1957,  lateral  view  of  internal  mould 
of  incomplete  cephalon,  same  horizon  as  8,  Skjaerholmen,  Oslo,  x 4-);  also  figured  by  Stormer  (1945,  pi.  1,  fig. 
10).  10,  PMO  100720,  ventral  view  of  cast  of  lower  lamella,  pygidium,  and  thorax,  top  of  Husbergoya  Shale 
Formation,  Rambergoya,  Oslo,  x 2\.  11,14,  morph  C,  PMO  103963,  frontal  and  anterolateral  views  of  cast 
of  cephalon,  same  horizon  as  8,  Hovedoya,  Oslo,  x 3.  13,  morph  C,  PMO  103964,  posterior  view  of  cast  of 

crushed  cephalon  and  incomplete  thorax,  note  weak  reticulation  on  posteromesial  parts  of  genal  lobe,  upper  2 
m of  Husbergoya  Shale  Formation,  Husbergoya,  Oslo,  x 4. 

Figs.  12,  15.  Tretaspis  askerensis  sp.  nov.  12,  PM064649,  frontal  view  of  cast  of  crushed  cranidium,  middle  part 
of  Grina  Shale  Member  of  the  Lunner  Formation,  Grina,  Hadeland,  x 4;  also  figured  by  Stormer  (1945,  pi.  1, 
fig.  1).  15,  PM06376,  posterolateral  view  of  cast  of  incomplete  cephalon,  from  either  the  lower  part  of  the 
Langara  Limestone-Shale  Formation  or  the  Husbergoya  Shale  Formation,  Hvalstad,  Asker,  x 6. 


PLATE  92 


OWEN,  trilobite  Tretaspis 


740  PALAEONTOLOGY,  VOLUME  23 

Girvan  are  of  this  type  and  are  very  similar,  if  not  identical,  to  the  Norwegian  morph  C.  Thus  T. 
sortita  broeggeri  differs  from  the  Scottish  form  only  in  the  proportions  of  constituent  morphs. 

Price  (1977,  pp.  784-785,  pi.  103,  figs.  1-7;  text-fig.  2)  assigned  material  to  T.  sortita  from  late 
Ashgill  mudstones  in  the  Meiford  area  and  commented  on  other  Welsh  material  probably  belonging 
to  this  species.  The  specimens  which  he  described  have  E2  developed  and  only  one  out  of  seven  lacks 
pits  in  I5.  Unlike  both  the  Norwegian  and  Scottish  forms,  the  genal  lobes  are  completely  smooth,  arcs 
Ij-Ej  share  short  sulci  in  only  a few  specimens  and  lists  are  less  well  developed. 

Dr.  Ingham  informs  me  (pers.  comm.  1976)  that  one  specimen  of  Tretaspis  from  the  type  unit  and 
locality  of  T.  latilimbus  latilimbus,  the  Upper  Johnstorp  Formation  in  Vastergotland,  has  a short  I5 
developed  (four  pits).  The  E pits  are  not  preserved  but  the  specimen  may  well  be  of  T.  sortita  type  and 
further,  well-localized,  collections  may  enable  greater  correlation  between  the  Swedish,  Norwegian, 
and  British  upper  Ashgill  sequences. 

Tretaspis  askerensis  sp.  nov. 

Plate  92,  figs.  12,  15;  Plate  93,  figs.  1-5. 

1902  Trinucleus  Wahlenbergi  Rouault;  Kiaer,  p.  78. 

1945  Tretaspis  seticornis  (Hisinger)  forma  typica;  Stormer,  p.  406,  pi.  1,  fig.  6. 

1978  Tretaspis  aff.  seticornis  seticornis  (Hisinger);  Owen,  p.  15. 

Holotype.  A cranidium  (PMO 100657)  from  either  the  Husbergoya  Shale  Formation  or  the  lower  part  of  the 
Langara  Limestone-Shale  Formation  (i.e.  ‘5a’  of  Brenchley  and  Newall  1975)  Holmenskjaeret,  Holmen,  Asker. 

Material,  localities,  and  horizons.  Four  incomplete  cranidia  from  the  type  horizon  and  locality,  a cephalon 
probably  from  ‘5a’  at  0vre  Nes  badestrand,  a cranidium  possibly  from  this  unit  at  Hvalstad,  three  cranidial 
fragments  from  2-3  m above  the  base  of  the  Husbergoya  Formation  on  Bronnoya  and  1 -4  m above  the  base  of 
this  unit  on  Langara,  all  Asker.  Two  cranidia  from  the  lowest  13  m of  the  Husbergoya  Formation  on  Kalvoya, 
Baerum.  Three  fragmentary  cranidia  from  a channel  conglomerate  in  the  upper  part  of  the  Langara  Formation 
on  Ostoya,  Baerum  (indicating  transport  from  the  west),  and  one  external  mould  of  a cephalon  from  the  Grina 
Shale  Member  of  the  Lunner  Formation  at  Grina  Hadeland. 

Diagnosis.  Pseudofrontal  lobe  strongly  swollen,  almost  circular  in  dorsal  view.  Arcs  Ex,  Ix,  I2,  and  In  complete.  A 
short  I3  is  present  in  a few  specimens.  f-Ej  sulci  deep.  E2  short  or  absent. 

Description.  Most  specimens  of  this  form  are  noticeably  smaller  than  those  of  other  Norwegian  species  but  the 
material  is  too  incomplete  to  quantify  this  adequately.  Pseudofrontal  lobe  strongly  swollen,  almost  circular  in 
dorsal  view  but  otherwise  the  proportions  of  the  glabella  and  genae  very  similar  to  those  of  T.  ceriodes  angelini. 
The  external  surface  of  the  glabella  and  genae  bears  a well-developed  reticulation  which  is  very  fine  except  on  the 
posteromesial  parts  of  the  genal  lobes  where  it  is  coarser.  Some  internal  moulds  fairly  strongly  reticulate.  Fringe 
steeply  declined.  Arcs  El5 11;  I2,  and  In  complete.  E2  developed  in  two  specimens  (of  four)  where  it  comprises  up 
to  six  pits.  I3  present  in  two  specimens  (of  seven)  where  it  contains  two  or  three  pits,  beginning  at  about  aR3. 
There  are  eight  pits  along  the  posterior  margin  of  the  fringe  in  three  specimens.  There  are  nineteen  pits  in  In  in  one 
topotype  specimen  and  the  Grina  Shale  cephalon  and  1 6^  in  a specimen  from  Kalvoya.  Arcs  E , and  I , share  deep 
sulci  over  all  but  the  posterior  part  of  the  fringe  and  are  out  of  phase  with  the  remaining  I arcs. 

Remainder  of  exoskeleton  unknown. 

Discussion.  The  fringe  development  of  T.  askerensis  resembles  that  of  T.  seticornis  and  the  Grina 
Shale  specimen  was  assigned  to  this  species  by  Stormer  (1945)  and  Owen  (1978).  T.  askerensis  differs 
in  its  deep  I j -Ej  sulci,  in  having  I3  developed  in  a few  specimens,  in  having  eight  (cf.  six  or  seven)  pits 
along  the  posterior  margin,  and  the  number  of  pits  in  In  extends  beyond  the  maximum  recorded  for  T. 
seticornis.  Clearly  the  very  limited  number  of  specimens  of  both  species  makes  objective  comparison 
very  difficult.  The  more  circular  outline  of  the  pseudofrontal  lobe  and  much  stronger  reticulation  also 
distinguish  the  younger  species  although  the  latter  character  may  have  little  taxonomic  value  (see 
Price  1977,  p.  781).  T.  seticornis  has  a very  short  stratigraphical  range,  being  restricted  to  low 
Pusgillian  strata  in  both  Norway  and  Sweden.  T.  askerensis  occurs  in  Rawtheyan  units  and  probably 


OWEN:  TRILOBITE  TRETASPIS 


741 


was  derived  from,  for  example,  T.  hadelandica  hadelandica  or  T.  latilimbus  norvegicus.  The  relatively 
small  size,  well-developed  reticulation  (see  Stormer  1930,  p.  65)  and  simple  fringe  morphology 
suggest  a neotenous  origin  for  the  species. 

Tretaspis  sagenosus  group? 

Tretaspis  kiaeri  Stormer,  1930 
Plate  93,  figs.  6-15;  text-fig.  9 


1921  Trinucleus;  Kiaer,  p.  500. 

1930  Tretaspis  kiaeri  Stormer,  pp.  50-55,  pi.  10,  figs.  1-6;  pi.  11,  fig.  12;  pi.  13,  fig.  13;  pi.  14,  figs.  1-3; 
text-figs.  21c,  23-26,  38. 

1945  Tretaspis  kiaeri  Stormer;  Stormer  (pars),  p.  403,  pi.  1,  fig.  12;  non  pp.  387,  406,  pi.  1,  fig.  1 1 [=  T. 
hadelandica  hadelandica ]. 

1953  Tretaspis  kiaeri;  Stermer,  p.  87. 

1959  Tretaspis  kiaeri  Stormer;  Harrington  in  Moore,  text-fig.  70c. 
non  1966  Tretaspis  kiaeri  Stormer;  Whittington,  pp.  90-92,  pi.  28,  figs.  1,6-12,  14. 
non  1968  Tretaspis  kiaeri  Stormer;  Whittington,  p.  93,  pi.  29,  figs.  1,  2,  4. 

1975  T.  kiaeri  Stormer;  Hughes  et  al .,  p.  563. 
non  1975  T.  aff.  kiaeri;  Hughes  et  al.,  p.  563. 

1979  Tretaspis  kiaeri  Stormer;  Owen,  pp.  250,  251,  252,  text-fig.  6. 

1979  Tretaspis  kiaeri  Stormer;  Bruton  and  Owen,  text-fig.  6. 

Holotype.  An  almost  complete  internal  mould  of  a cephalon  (PMO  H197)  from  the  Hogberg  Member  of  the 
Solvang  Formation,  Frognoya,  Ringerike. 

Material,  locality,  and  horizon.  The  species  is  known  only  from  the  type  horizon  and  locality  from  which  many 
hundreds  of  disarticulated  skeletal  elements  are  known. 

Description.  Proportions  of  glabella  and  genal  lobes  similar  to  those  of  T.  ceriodes  angelini  except  that  the 
glabella  is  a little  more  inflated  and  overhangs  the  fringe  a little.  Reticulation  variable.  On  the  glabella  it  is 
coarsest  around  the  median  node  and  extends  to  a transverse  line  at  the  maximum  width  (tr.)  of  the  occiput.  The 
reticulation  of  the  genal  lobes  is  finer  and  more  subdued  than  that  of  the  mesial  part  of  the  glabella  and  becomes 
finer  abaxially.  On  some  internal  moulds  there  is  a faint  reticulation  on  the  genal  lobes  and,  less  commonly,  the 
glabella.  Fringe  steeply  declined  laterally,  less  so  across  In  mesially,  in  front  of  which  it  is  vertical.  Arcs  Ej_2, 1 ,_3, 
and  In  complete  mesially  and  posteriorly.  I4  is  continuous  mesially  but  extends  to  the  posterior  margin  in  only  5% 
of  sixty-one  specimens.  Two  morphs  are  defined  on  the  absence  (A)  or  presence  (B)  of  I5  which  occurs  in  35%  of 
eighty-one  specimens  and  extends  mesially  in  22%  of  the  twenty-three  specimens  in  which  its  frontal  extent  can 
be  determined  (Table  1).  The  range  of  variation  in  selected  fringe  characters  is  shown  on  text-fig.  9.  Two  sets  of 
radii  are  developed  and  pits  in  the  outer  set,  Il5  E^,  share  sulci  to  the  posterior  part  of  the  fringe  in  some 
specimens  but  in  a few  this  sulcation  is  less  extensive  and  L becomes  discrete  as  far  forwards  as  bR5.  Very  fine 
lists  are  developed  between  all  I arcs  except  I4  and  I5. 

Hypostoma  unknown.  Thorax  similar  to  that  of  T.  seticornis,  although  it  is  not  known  whether  or  not  median 
tubercles  are  present.  The  pygidial  rachis  commonly  has  up  to  six  transversely  directed  furrows  bearing  deep 
apodemal  pits  distally.  These  furrows  are  progressively  less  well  incised  rearwards  along  the  rachis  and  on  well- 
preserved  specimens  (PI.  93,  fig.  9;  Stormer  1930,  pi.  10,  fig.  4)  a further  three  to  five  pairs  of  apodemal  markings 
are  seen,  the  posterior  two  or  three  pairs  being  situated  on  the  anterior  part  of  the  border.  Three  pairs  of  weakly 
developed  pleural  ribs  present. 

Discussion.  Whittington  (1966,  1968)  ascribed  specimens  from  the  Ashgill  of  Wales  to  T.  kiaeri.  One 
of  these  (1966,  pi.  28,  fig.  13)  was  referred  to  Nankinolithus  Lu  by  Hughes  et  al.  (1975,  p.  559).  The 
remainder  comprise  at  least  three  distinct  forms  of  Tretaspis  and  have  been  reassessed  by  Price  ( 1 977, 
pp.  786-787)  who  considered  specimens  from  the  Rhiwlas  Limestone  (probably  Rawtheyan)  figured 
by  Whittington  (1968,  pi.  28,  figs.  12,  16)  to  be  similar  to  T.  calcaria  Dean,  1971,  a form  described 
originally  from  the  Chair  of  Kildare  Limestone  (probably  Rawtheyan)  in  Eire.  T.  calcaria  is  almost 
certainly  related  to  T.  kiaeri  but  differs  in  having  I4  always  complete  posteriorly,  I5  more  extensive, 
and  all  complete  arcs  have  a higher  pit  count  (e.g.  30-31  cf.  20^-27^  in  Ex).  As  noted  by  Price,  the 


742 


PALAEONTOLOGY,  VOLUME  23 


poorly  preserved  Rhiwlas  Limestone  material  is  difficult  to  compare  with  Dean’s  species  but 
differences  in  fringe  pitting  seem  slight. 

Other  British  and  Irish  forms  previously  assigned  to  T.  kiaeri  have  been  reassessed  by  Ingham 
(1970,  pp.  44-57)  and  Price  (1974,  pp.  844-847;  1977,  pp.  766-778).  Most  are  clearly  members  of  the 
T.  moeldenensis  group  and  thus  are  distinguished  from  T.  kiaeri  primarily  in  having  complete  radial 
alignment  of  the  fringe  pits.  A few  are  T.  seticornis  group  members  and  have  E2  incomplete  mesially. 


EVOLUTION  OF  THE  TRETASPIS  SETICORNIS  GROUP 


The  study  of  populations  of  Tretaspis  from  Norway  indicates  that  the  phylogenetic  relationships  are 
more  complex  than  was  thought  previously  and  that  a purely  typological  approach  to  their  taxonomy 
is  not  possible.  Nevertheless,  the  broad  evolutionary  history  of  the  T.  seticornis  group  is  becoming 
clear  (text-fig.  1). 


n 1 1 m n rhn  fi 

- 3 4 5 6 7 

Pits  in  I5 


text-fig.  9.  Histograms  showing  the  range  of  variation  in  fringe  features  seen 
in  all  available  specimens  of  Tretaspis  kiaeri  with  a comparison  of  the  range, 
mean,  and  sample  standard  deviation  of  the  two  morphs  (A  and  B)  present  in 
the  species. 


EXPLANATION  OF  PLATE  93 

Figs.  1-5.  Tretaspis  askerensis  sp.  nov.  1-3,  holotype,  PM0100657,  dorsal,  lateral,  and  anterolateral  views  of 
partially  exfoliated  cranidium,  Husbergoya  Shale  Formation,  or  lower  part  of  Langara  Limestone-Shale 
Formation,  Holmenskjaeret,  Holmen,  Asker,  x 7.  4,  PMO80463,  anterolateral  view  of  partially  exfoliated 
cranidium,  same  horizon  and  locality  as  1-3,  x 10.  5,  PM0100878,  cast  of  flattened  incomplete  cephalon, 
probably  from  the  type  unit,  0vre  Nes  badestrand,  Nesbru,  Asker,  x 6. 

Figs.  6-15.  Tretaspis  kiaeri  Stormer,  Hogberg  Member  of  the  Solvang  Formation,  Frognoya,  Ringerike.  6,  10, 
holotype,  morph  B,  PMO  HI  97,  dorsal  and  frontal  views  of  internal  mould  of  cephalon,  x 3£;  also  figured  by 
Stormer  (1930,  pi.  10,  fig.  1).  7,  morph  B,  PMO  H338,  lateral  view  of  internal  mould  of  cephalon,  x 3;  also 
figured  by  Stormer  (1930,  pi.  10,  fig.  3).  8,  PMO103965,  dorsal  view  of  cast  of  pygidium  and  incomplete 
thorax,  x4^.  9,  PMO103966,  dorsal  view  of  internal  mould  of  pygidium,  x4J.  1 1,  morph  A,  PMO  103967, 
anterolateral  view  of  internal  mould  of  incomplete  cephalon,  x 2.  12,  morph  A,  PMO  H208,  posterolateral 
view  of  incomplete  partially  exfoliated  cranidium  showing  pitting  along  the  marginal  band,  x 4;  also  figured 
by  Stormer  (1930,  pi.  11,  fig.  12).  13,  PMO  103968,  slightly  oblique  dorsal  view  of  cast  of  glabella  and  left 
genal  lobe,  note  glabellar  reticulation,  x 14,  morph  B,  PMO  103969,  dorsal  view  of  internal  mould  of 
cephalon  and  part  of  thorax,  same  specimen  as  8,  x 4.  1 5,  morph  B,  PM0354,  oblique  anterolateral  view  of 
cephalon,  x 3;  also  figured  by  Stormer  (1945,  pi.  1,  fig.  12). 


PLATE  93 


OWEN,  trilobite  Tretaspis 


744 


PALAEONTOLOGY,  VOLUME  23 


The  earliest  known  species  of  Tretaspis  from  the  Anglo-Welsh  and  Scandinavian  areas  is  T. 
ceriodes  which  is  restricted  to  latest  Caradoc  units  in  all  these  areas.  The  species  is  polymorphic  in 
Norway  and  almost  certainly  gave  rise  to  the  T.  seticornis  group,  the  replacement  of  the  former  by  the 
latter  being  geologically  instantaneous  and  an  excellent  tool  in  recognizing  the  Caradoc-Ashgill 
boundary  (Owen  1979,  p.  251).  The  earliest  representatives  of  this  group  are  distinct  in  different  areas 
with  T.  hadelandica  in  England  and  Hadeland  and  T.  seticornis  in  Oslo-Asker,  Ringerike,  and 
Sweden.  This  rapid  speciation  involved  the  development  of  two  sets  of  pit  radii  and,  with  the 
exception  of  some  members  of  early  T.  hadelandica  populations,  the  restriction  of  E2  to  the  lateral 
parts  of  the  fringe.  The  polymorphic  nature  of  the  ancestral  T.  ceriodes  populations  accounts  for  all 
other  fringe  features  of  the  early  T.  seticornis  group  forms.  Local  populations  of  T.  hadelandica 
became  isolated  very  early  on,  giving  rise  to  what  are  interpreted  as  geographical  subspecies.  The  T. 
moeldenensis  group  persisted  into  the  Ashgill  in  Britain  but  not  in  Scandinavia. 

In  Britain,  T.  hadelandica  is  now  interpreted  as  ranging  from  earliest  Pusgillian  to  mid/late 
Rawtheyan  with  a series  of  stratigraphical  subspecies  showing  a progressive  simplification  of  fringe 
characters  ( T . h.  convergens — T.  h.  deliquus — T.  h.  brachystichus)  followed  by  a slight  reversal  of  this 
trend  within  T.  h.  brachystichus  which  may  have  been  continued  with  the  development  of  T. 
Tatilimbus'  distichus.  This  reinterpretation  strengthens  the  stratigraphical  usefulness  of  the  British 
forms  especially  in  view  of  the  long-ranging  homeomorphs  present  in  Norway.  The  origins  of  the 
Irish  and  Scottish  T.  sp.  from  which  T.  persulcatus  were  descended  are  unclear. 

In  Hadeland,  T.  hadelandica  hadelandica  persisted  from  early  Pusgillian  to  Rawtheyan  times,  and 
although  there  are  differences  in  the  percentages  of  constituent  morphs  in  different  units,  these  are 
considered  to  reflect  ecological  rather  than  temporal  controls.  T.  hadelandica  may  have  given  rise  to  a 
homeomorph  of  T.  seticornis,  T.  askerensis  which  occurs  in  Hadeland  and  Asker. 

In  Oslo-Asker,  Ringerike,  and  Sweden,  T.  seticornis  has  a short  stratigraphical  range  and  gave  rise 
to  another  short  ranging  form,  T.  hisingeri.  In  Ringerike,  T.  seticornis  also  gave  rise  to  T.  anderssoni, 
a form  which  has  a very  narrow  range  of  variation  throughout  its  range  from  mid-Pusgillian  to  early 
Rawtheyan.  In  Oslo-Asker,  T.  hisingeri  is  replaced  by  T.  latilimbus  norvegicus,  a polymorphic  form  of 
uncertain  origin  which  extends  well  into  the  Rawtheyan  and  which  almost  certainly  gave  rise  to 
T.  sortita  broeggeri.  One  of  the  morphs  constituting  T.  latilimbus  norvegicus  is  by  far  the  dominant 
form  in  the  nominate  subspecies  which  is  a Swedish  taxon  developed  during  the  Rawtheyan. 
Populations  of  T.  sortita  sortita  from  the  late  Rawtheyan  of  Scotland  differ  from  T.  sortita  broeggeri 
in  the  proportions  of  constituent  morphs. 

There  is  still  very  little  information  on  bed-by-bed  changes  in  populations  of  Tretaspis,  and  the 
Norwegian  material  is  not  sufficiently  abundant  for  such  a study.  There  is  a suggestion  that  the 
development  of  phenotypes  in  T.  ceriodes  angelini  is  to  some  extent  progressive  but  as  far  as  morphs 
B,  C,  and  D are  concerned  this  represents  no  more  than  an  increase  in  the  upper  limit  of  the  range  of 
variation.  Many  forms  have  long  stratigraphical  ranges  within  which  there  is  no  directional  change. 
The  only  likely  example  of  evolutionary  trends  are  the  zigzag  evolution  seen  in  the  British  T. 
hadelandica  subspecies  and  the  introduction  of  a third  morph  to  produce  T.  sortita  broeggeri  from  T. 
latilimbus  norvegicus.  The  latter  change  was  fairly  abrupt  as  was  the  development  of  the  T.  seticornis 
group  itself.  There  is  insufficient  evidence  to  say  whether  or  not  the  changes  in  the  British  subspecies 
of  T.  hadelandica  are  gradual.  Neoteny  is  thought  to  have  produced  two  species,  T.  hisingeri  and  T. 
askerensis  and  probably  also  T.  ceriodes  from  the  T.  sagenosus  group. 

Acknowledgements.  I am  very  grateful  to  Dr.  J.  K.  Ingham  for  his  considerable  help  and  encouragement  and  for 
his  comments  on  an  earlier  draft  of  this  paper.  I have  also  benefited  from  discussions  with  Professor  H.  B. 
Whittington  and  the  late  Professor  L.  Stormer.  I thank  Mr.  A.  Buxton  and  Mr.  J.  Smith  for  their  help  in 
preparing  the  figures  and  plates,  Drs.  D.  L.  Bruton  (Paleontologisk  Museum,  Oslo),  R.  A.  Fortey  (British 
Museum  (Natural  History)),  and  V.  Jaanusson  (Riksmuseum,  Stockholm)  for  access  to  collections  in  their  care, 
and  Dr.  P.  J.  Brenchley  and  his  group  for  showing  me  their  collections  (now  PMO).  Most  of  the  work  was  carried 
out  during  the  tenures  of  a N.E.R.C.  studentship  at  Glasgow  University  and  a N.A.T.O.  fellowship  at  the 
Paleontologisk  Museum,  Oslo. 


OWEN:  TRILOBITE  TRET  A S P IS 


745 


REFERENCES 

andersson,  J.  G.  1894.  Ueber  Blocke  aus  dem  jungeren  Untersilur  auf  der  Insel  Oland  varkommend.  Ofv.  Kongl. 
Vet.-Akad.  Fork.  1893,  8,  521-540. 

angelin,  N.  p.  1854.  Palaeontologica  Scandinavica  I:  Crustacea  formationis  transitionis,  Fasc.  2,  21-92,  pis. 
25-41. 

asklund,  b.  1936.  Die  Fauna  in  einem  Geschiebe  aus  der  Trinucleusstufe  in  Jamtland.  Sver.  geol.  Unders.  Afh. 
(C).  400,  1-6,  pis.  1-2. 

bassler,  R.  s.  1915.  Bibliographic  index  of  American  Ordovician  and  Silurian  fossils.  Bull.  U.S.  natn.  Mus.  92, 
1-1521,  pis.  1-4. 

begg,  J.  L.  1944.  On  the  fringe  of  Tretaspis.  Geol.  Mag.  81,  113-117,  pi.  5. 

bolton,  T.  E.  1970.  Subsurface  Ordovician  fauna,  Anticosti  Island,  Quebec.  Bull.  geol.  Surv.  Can.  187,  31-39, 
136-139,  pis.  6,  7. 

brenchley,  p.  J.  and  newall,  G.  1975.  The  stratigraphy  of  the  Upper  Ordovician  Stage  5 in  the  Oslo-Asker 
district,  Norway.  Norsk  geol.  Tidsskr.  55,  243-275. 
brogger,  w.  c.  1887.  Geologisk  kart  over  oerne  ved  Kristiania.  Nyt.  Mag.  f.  Naturvid.  31,  1-36. 
bruton,  D.  L.  and  owen,  A.  w.  1 979.  Late  Caradoc -early  Ashgill  trilobite  distribution  in  the  central  Oslo  Region, 
Norway.  Norsk  geol.  Tidsskr.  59,  213-222. 

cave,  R.  1960.  A new  species  of  Tretaspis  from  South  Wales.  Geol.  Mag.  97,  334-337,  pi.  10. 

— 1965.  The  Nod  Glas  sediments  of  Caradoc  age  in  North  Wales.  Geol.  J.  4,  279  298,  pi.  12. 

cech,  s.  1975.  Cranidial  reticulation  and  functional  morphology  of  the  cephalic  fringe  in  Trinucleidae 
(Trilobita).  Vest.  Ustr.  ust.geol.  50,  173-177,  pis.  1-4. 

cocks,  L.  R.  M.  and  price,  d.  1975.  The  biostratigraphy  of  the  upper  Ordovician  and  lower  Silurian  of  south-west 
Dyfed,  with  comments  on  the  Hirnantia  fauna.  Palaeontology,  18,  703-724,  pis.  81-84. 
dean,  w.  t.  1959.  The  stratigraphy  of  the  Caradoc  Series  in  the  Cross  Fell  Inlier.  Proc.  Yorks,  geol.  Soc.  32, 
185-227. 

— 1961 . Trinucleid  trilobites  from  the  higher  Dufton  Shales  of  the  Caradoc  Series  in  the  Cross  Fell  Inlier.  Ibid. 
33,  119-134,  pis.  7-9. 

— 1962.  The  trilobites  of  the  Caradoc  Series  in  the  Cross  Fell  Inlier  of  northern  England.  Bull.  Br.  Mus.  nat. 
Hist.  {Geol.),  7,  65-134,  pis.  6-18. 

— 1963.  The  Ordovician  trilobite  faunas  of  South  Shropshire,  IV.  Ibid.  9,  1-18,  pis.  1,  2. 

— 1971.  The  trilobites  of  the  Chair  of  Kildare  Limestone  (Upper  Ordovician)  of  eastern  Ireland.  Palaeontogr. 
Soc.  [ Monogr .],  1,  1-60,  pis.  1-25. 

fearnsides,  w.  G.,  elles,  G.  and  smith,  b.  1907.  The  Lower  Palaeozoic  rocks  of  Pomeroy.  Proc.  R.  Ir.  Acad.  26B, 
97-128,  pis.  7,  8. 

havlicek,  v.  and  vanek,  J.  1966.  The  biostratigraphy  of  the  Ordovician  of  Bohemia.  Shorn.  Geol.  ved.  [P],  8, 
7-69,  pis.  1-16. 

hawle,  I.  and  corda,  A.  J.  c.  1847.  Prodrom  einer  Monographie  der  bohmischen  Trilobiten.  176  pp.  Prague. 
hayami,  i.  and  ozawa,  T.  1975.  Evolutionary  models  of  lineage  zones.  Lethaia,  8,  1-14. 
hisinger,  w.  1840.  Lethaea  Svecica  seu  Petrificata  Sveciae,  iconibus  et  characteribus  illustrata.  Suppl. 
secundum.  Holmiae. 

HOLTEDAHL,  o.  and  schetelig,  J.  1923.  Kartbladet  Gran.  Norges  geol.  Unders.  97,  1-46. 
hughes,  c.  p.  1970.  Statistical  analysis  and  presentation  of  trinucleid  (Trilobita)  fringe  data.  Palaeontology,  13, 
1-9. 

— ingham,  j.  k.  and  addison,  R.  1975.  The  morphology,  classification  and  evolution  of  the  Trinucleidae 
(Trilobita).  Phil.  Trans.  R.  Soc.  Lond.  B272,  537-607,  pis.  1-10. 

ingham,  j.  k.  1970.  A monograph  of  the  Upper  Ordovician  trilobites  from  the  Cautley  and  Dent  districts  of 
Westmorland  and  Yorkshire.  Palaeontogr.  Soc.  [Monogr],  1,  1-58,  pis.  1-9. 

— 1978.  Geology  of  a continental  margin  2:  middle  and  late  Ordovician  transgression,  Girvan. 

In  bowes,  D.  R.  and  leake,  b.  e.  (eds.).  Crustal  evolution  in  northwestern  Britain  and  adjacent  regions.  Geol. 
J.  Special  Issue,  10,  163-176. 

jaanusson,  v.  1956.  Undersuchungen  uber  der  oberordovizischen  Lykholm-Stufenkomplex  in  Estland.  Bull, 
geol.  Instn.  Univ.  Uppsala,  36,  369-400,  pi.  1 . 

— and  martna,  J.  1948.  A section  from  the  Upper  Chasmops  series  to  the  Lower  Tretaspis  series  at  Fjacka 
rivulet  in  the  Siljan  Area,  Dalarne.  Ibid.  32,  183-193. 

kiaer,  j.  1897.  Faunistische  Uebersicht  der  Etage  5 des  norwegischen  Silursystem.  Skr.  Norsk  Vidensk.-Akad. 
Mat.  Naturv.  Kl.  3,  1 -76. 


746 


PALAEONTOLOGY,  VOLUME  23 


kiaer,  j.  1902.  Etage  5 i Asker  ved  Kristiania.  Norges  geol.  Under s.  34,  1-112. 

— 1921.  En  ny  zone  i Norges  midtre  Ordovicium.  Geol.  Foren.  Stockh.  Forh.  43,  499-502. 
kielan,  z.  1957.  On  the  trilobite  family  Staurocephalidae.  Acta.  Pal.  Polon.  2,  155-182. 

— 1960.  Upper  Ordovician  trilobites  from  Poland  and  some  related  forms  from  Bohemia  and  Scandinavia. 
Palaeont.  Pol.  11,  i-vi,  1-198,  pis.  1-36. 

lamont,  A.  1935.  The  Drummuck  Group,  Girvan;  A stratigraphical  revision  with  descriptions  of  new  fossils 
from  the  lower  part  of  the  group.  Trans,  geol.  Soc.  Glasg.  19,  288-332,  pis.  7-9. 

1941.  Trinucleidae  in  Eire.  Ann.  Mag.  nat.  Hist.  (1 1),  8,  438-469,  pi.  5. 

lauritzen,  0.  1973.  The  Middle  Ordovician  of  the  Oslo  Region,  Norway.  24.  Stage  4b  at  Lunner,  Hadeland. 
Norsk  geol.  Tidsskr.  53,  25-40. 

lesperance,  p.  j.  1968.  Trilobite  faunas  of  the  White  Head  Formation,  Perce  Region,  Quebec.  J.  Paleont.  42, 
811-826. 

— and  bertrand,  R.  1976.  Population  systematics  of  the  Middle  and  Upper  Ordovician  trilobite  Cryptolithus 
from  the  St.  Lawrence  Lowlands  and  adjacent  areas  of  Quebec.  Ibid.  50,  598-613. 

linnarsson,  j.  G.  o.  1869.  Om  Vestergotlands  Cambriska  och  Siluriska  Aflagringar.  K.  svenska  Vetensk-Akad. 
Hand/.  8(2),  1-89,  pis.  1,2. 

loven,  s.  L.  1845.  Svenska  Trilobiter.  Ofver.  Kgl.  Vet.-Akad.  Forhandl.  andra  argangen  (1845),  46-56,  104-1 11, 
pis.  1,  2. 

mccoy,  f.  1 849.  On  the  classification  of  some  British  fossil  Crustacea,  with  notices  of  new  forms  in  the  University 
Collection  at  Cambridge.  Ann.  Mag.  nat.  Hist.  (2),  4,  161-179,  330-335,  392-414. 
mcnamara,  k.  j.  1979.  Trilobites  from  the  Coniston  Limestone  Group  (Ashgill  Series)  of  the  Lake  District, 
England.  Palaeontology,  22,  53-92,  pis.  7-12. 
mayr,  e.  1969.  Principles  of  systematic  zoology,  x,  428  pp.  McGraw  Hill,  New  York. 

miller,  J.  1976.  The  sensory  fields  and  life  mode  of  Phacops  rana  (Green,  1832)  (Trilobita).  Trans.  R.  Soc.  Edinb. 
69,  337-367,  pis.  1-4. 

moore,  r.  c.  (ed.).  1959.  Treatise  on  invertebrate  palaeontology.  Part  O,  Arthropoda  1.  xix,  560  pp.  Geol.  Soc. 
Amer.  and  Univ.  Kansas  Press,  Lawrence. 

owen,  A.  w.  1977.  Upper  Ordovician  stratigraphy  and  trilobite  faunas  of  the  Oslo  region,  with  special  reference 
to  Hadeland  and  Ringerike.  Unpubl.  Ph.D.  thesis,  Univ.  of  Glasgow. 

— 1978.  The  Ordovician  and  Silurian  stratigraphy  of  Central  Hadeland,  South  Norway.  Norges  geol.  Unders. 
338,  1-23,  pi.  1. 

— 1979  (for  1978).  The  upper  Ordovician  succession  at  Norderhov  and  on  Frognoya  in  Ringerike,  Norway. 
Norsk  geol.  Tidsskr.  58,  245-258. 

— 1980.  A new  species  of  Cryptolithus  (Trilobita)  from  the  late  Ordovician  of  Norway.  J.  Paleont.  54, 
144-148. 

portlock,  J.  E.  1843.  Report  on  the  geology  of  the  county  of  Londonderry,  and  of  parts  of  Tyrone  and  Fermanagh. 
xxxi,  784  pp.,  pis.  1-38,  A-l.  Dublin  and  London. 

pribyl,  A.  and  vanek,  J.  1969.  Trilobites  of  the  family  Trinucleidae  Hawle  et  Corda,  1847  from  the  Ordovician  of 
Bohemia.  Sb.  geol.  ved.  Paleontologie,  11,  85-137. 

price,  d.  1973.  The  age  and  stratigraphy  of  the  Sholeshook  Limestone  of  South-west  Wales.  Geol.  J.  8, 225-246. 

— 1974.  Trilobites  from  the  Sholeshook  Limestone  (Ashgill)  of  South  Wales.  Palaeontology,  17, 841-868,  pis. 
112-116. 

— 1977.  Species  of  Tretaspis  (Trilobita)  from  the  Ashgill  Series  in  Wales.  Ibid.  20,  763-792,  pis.  98-103. 
reed,  F.  r.  c.  1935.  The  Lower  Palaeozoic  trilobites  of  Girvan.  Supplement  No.  3.  Palaeontogr . Soc.  [ Monogr .], 

1-64,  pis.  1-4. 

Schmidt,  f.  1894.  Revision  der  Ostbaltischen  Silurischen  Trilobiten  Abth.  IV.  Mem.  Acad.  imp.  Sci.  St.  Petersb. 
(7),  42,  (5),  1-94,  pis.  1-6. 

schuchert,  c.  and  cooper,  G.  A.  1 930.  Upper  Ordovician  and  Lower  Devonian  stratigraphy  and  palaeontology 
of  Perce,  Quebec.  Part  II.  New  Species  from  the  Upper  Ordovician  of  Perce.  Am.  J.  Sci.  (5),  20,  265-288, 
365-392,  pis.  1-5. 

stgrmer,  l.  1930.  Scandinavian  Trinucleidae  with  special  reference  to  Norwegian  species  and  varieties.  Skr. 
Norsk  Vidensk.-Akad.  Mat.  Naturv.  Kl.  4,  1-111,  pis.  1-14. 

— 1934.  Cambro-Silurian  zones  of  the  Oslo  Region,  with  a brief  correlation  between  British  and  Norwegian 
sections.  In  holtedahl,  o.  et  al.  1934.  The  geology  of  parts  of  Southern  Norway.  Proc.  geol.  Ass.  45, 307-377, 
pis.  22-32. 

— 1945.  Remarks  on  theTretaspis  (Trinucleus)  Shales  of  Hadeland  with  description  of  trilobite  faunas.  Norsk 
geol.  Tidsskr.  25,  379-425,  pis.  1-4. 


OWEN:  TRILOBITE  TRETASPIS 


747 


stormer,  l.  1953.  The  Middle  Ordovician  of  the  Oslo  Region,  Norway.  1 . Introduction  to  stratigraphy.  Ibid.  31, 
37-141,  pis.  1-6. 

strand,  T.  and  henningsmoen,  G.  1960.  Cambro-Silurian  stratigraphy.  In  holtedahl,  o.  (ed.)  Geology  of 
Norway.  Norges  geol.  Unders.  208,  128-169,  pis.  7,  8. 

tomczyk,  h.  1962.  Stratigraphy  of  Old  Palaeozoic  sediments  from  bore-holes  at  Vszkowce  near  Lubaczow.  In 
passendorfer,  e.  (ed.)  Ksiega  Dam.  Prof.  J.  Samsomowicza.  Polska  Akad.  Nauk.  Geol.  123-141  [Polish], 
141-148  [Russian  and  English  summaries],  pis.  25-29. 

tornquist,  s.  L.  1883.  Ofversicht  ofver  Bergbyggnaden  inom  Siljansomradet.  Sver.  geol.  Unders.  Afh.  (C),  57, 
1-59. 

— 1884.  Undersokninger  ofver  Siljansomradets  trilobitfauna.  Ibid.  (C),  66,  1-101,  pis.  1-3. 

Whittington,  h.  b.  1941.  Silicified  Trenton  trilobites.  J.  Paleont.  15,  492-522,  pis.  72-75. 

— 1966.  A Monograph  of  the  Ordovician  trilobites  of  the  Bala  area,  Merioneth.  Palaeontogr.  Soc.  [ Monogr .], 
3,  63-92,  pis.  19-28. 

— 1968.  A Monograph  of  the  Ordovician  trilobites  of  the  Bala  area,  Merioneth.  Ibid.  4,  93-138,  pis.  29-32. 

A.  w.  OWEN 

Department  of  Geology 

Typescript  received  22  August  1979  The  University 

Revised  typescript  received  7 January  1980  Dundee  DD1  4HN 


A TECHNIQUE  FOR  REVEALING  THE  STEREOM 
STRUCTURE  OF  FOSSIL  CRINOIDS 

by  G.  d.  sevastopulo  and  j.  b.  keegan 


Abstract.  The  stereom  of  fossil  crinoid  ossicles  preserved  in  an  argillaceous  matrix  can  be  revealed  by  treating 
them  with  hydrofluoric  acid.  The  clay  filling  the  stereom  pores  is  dissolved  and  the  skeletal  calcite  is  faithfully 
replaced  by  fluorite.  Features  discovered  in  selected  Lower  Carboniferous  crinoid  ossicles  prepared  by  this 
method  include  the  following:  large  canals  penetrating  the  areola  in  the  columnals  of  a particular  inadunate 
crinoid;  triple  aboral  nerve  canals,  and  labyrinthic  stereom  in  the  muscle  fossae  of  distinctive  inadunate 
brachials;  and  a regular  arrangement  of  trabeculae  forming  a cubic  structure  in  the  stereom  of  flexible  crinoid 
brachials. 

Over  the  last  decade,  there  have  been  a number  of  studies  of  the  detailed  morphology  of  recent 
echinoderm  ossicles,  using  scanning  electron  microscopy.  Of  these,  we  pick  out  the  surveys  of 
crinoid  microstructure  by  Macurda  and  Meyer  (1975,  1976)  and  the  extensive  investigations  of 
the  crinoid  stem  by  Roux  (1970,  1971,  1974«,  19746,  1975)  as  having  particular  significance  for  fossil 
crinoid  studies.  In  all  of  these  studies  the  architecture  of  the  stereom  has  been  shown  to  have  a 
functional  significance.  Unfortunately  the  details  of  the  stereom  are  difficult  to  discern  in  most  fossil 
material,  because  carbonate  cements  precipitated  epitaxially  on  the  skeletal  calcite  occlude  the 
stereom  pore  spaces.  In  some  examples,  however,  the  stereom  is  clearly  visible  in  thin  section.  The 
most  common  cases  of  this  are  when  the  stereom  pores  are  filled  by  an  iron-rich  carbonate  cement 
(which  can  be  differentiated  by  staining),  or  by  micritic  sediment  or  cement,  or  by  iron  sulphides  or 
oxides,  or  by  clay  minerals.  When  such  mineralogical  or  textural  differences  are  exploited  by  natural 
weathering  or  by  controlled  acid  etching,  the  three-dimensional  stereom  architecture  may  be 
revealed:  Lane  and  Macurda  (1975)  established  the  presence  of  muscular  articulation  in  naturally 
weathered  brachials  of  the  Pennsylvanian  cladid  crinoid,  Aesiocrinus\  and  Lapham,  Ausich,  and 
Lane  (1976)  have  illustrated  the  structure  of  the  stereom  of  Mississippian  crinoid  ossicles  which  had 
been  etched  in  weak  formic  acid. 

Whilst  trying  to  recover  miopores  from  a Carboniferous  marine  shale,  we  accidentally  discovered 
that  some  crinoid  ossicles  treated  with  hydrofluoric  acid  (HF)  showed  surprisingly  detailed 
microstructure:  clay  filling  the  stereom  pores  was  dissolved  and  the  calcite  of  the  ossicles  was 
faithfully  replaced  by  fluorite,  a process  which  has  been  named  fluoridization  by  Upshaw,  Todd,  and 
Allen  (1957,  p.  793).  The  use  of  hydrofluoric  acid  in  the  preparation  of  calcareous  fossils  has  been 
independently  (and  often  accidentally)  discovered  several  times  (Cookson  and  Singleton  1954; 
Grayson  1956;  Wetzel  1921).  Most  stress  has  been  laid  on  the  translucent  nature  of  fluoridized  fossils 
when  they  are  immersed  in  liquid.  Sohn  (1956)  was  able  to  make  visible  ostracode  muscle  scars  by 
treating  the  valves  with  HF,  and  Upshaw  et  al.  (1957)  illustrated  the  internal  structures  of  fluoridized 
foraminifers.  Sprinkle  and  Gutschick  (1967)  used  HF  to  prepare  blastoids  preserved  in  a fine-grained 
sandstone. 

We  have  applied  the  fluoridization  technique  to  Carboniferous  crinoid  material  preserved  in  a 
variety  of  rock  types,  ranging  from  plastic  clays  of  the  mid-western  United  States  to  indurated  silty 
mudstones  from  Ireland.  Examples  of  the  results  obtained  are  shown  in  Plates  94  and  95. 


(Palaeontology,  Vol.  23,  Part  4,  1980,  pp.  749-756,  pis.  94-95.| 


750 


PALAEONTOLOGY,  VOLUME  23 


METHODS 

The  material  for  which  the  fluoridization  technique  is  most  effective  is  that  preserved  in  clay  mudstones  or  shales 
where  the  clay  has  penetrated  deeply  into  the  stereom  pores.  We  have  generally  used  bulk  samples  rather  than 
attempting  to  fluoridize  particular  individual  specimens,  because  of  the  risk  of  damage.  However,  most  of  the 
microcrinoids  described  by  Lane  and  Sevastopulo  (in  press)  were  first  picked  from  washed  clays  and  then 
fluoridized. 

It  is  worth  while  to  remove  as  much  matrix  from  the  sample  as  possible.  Soft  clays  may  be  disaggregated  by 
being  air-dried,  soaked  in  paint  thinner  or  paraffin,  and  then  vigorously  boiled  in  water  with  soda  ash.  More 
indurated  mudstones  and  shales  may  require  simmering  in  Quaternary  ‘O’  (Zingula  1968),  but  since  that 
detergent  is  weakly  acid,  prolonged  treatment  results  in  some  etching  of  skeletal  calcite;  we  prefer  to  treat 
particularly  intractable  samples  directly  with  HF. 

The  partially  cleaned  fossil  material  is  reacted  with  HF;  the  optimum  strength  of  the  acid  and  length  of  the 
reaction  time  vary  from  sample  to  sample.  We  have  used  48%  HF  and  reaction  times  of  between  5 minutes  and 
1 hour  for  small  specimens;  for  larger  specimens  weaker  acid  (approximately  6%)  and  longer  reaction  times  (up 
to  24  hours)  as  advocated  by  Grayson  (1956,  p.  78)  lead  to  better  results.  The  fluoridization  can  be  judged  to  have 
proceeded  far  enough  when  the  surfaces  of  the  ossicles  appear  bleached;  it  is  not  necessary  to  convert  whole 
ossicles  to  fluorite. 

Two  adverse  effects  can  occur  during  fluoridization.  Firstly,  the  ossicles  may  crack  and  pieces  may  spall  off. 
This  can  be  largely  avoided  by  reducing  the  reaction  time  to  a minimum  and  by  diluting  the  acid.  Secondly, 
a glaze-like  precipitate  of  fluorite  may  form  on  the  surface  of  the  ossicles.  This  can  be  prevented  by  using  a large 
enough  quantity  of  acid  (we  have  found  five  times  the  volume  of  material  being  fluoridized  a suitable  amount). 
When  the  specimens  have  been  fluoridized,  they  should  be  thoroughly  washed  and  dried.  Specimens  for  study 
under  the  scanning  electron  microscope  should  be  transferred  to  stubs  immediately,  because  their  delicate 
surfaces  can  be  easily  damaged  by  abrasion. 

Although  we  have  been  interested  principally  in  the  preparation  of  crinoid  material,  our  bulk  samples  have 
contained  many  other  fossils,  most  of  which  appear  perfectly  preserved  after  fluoridization.  We  believe  that  the 
technique  may  have  general  application  in  cleaning  small  fossils  for  study  under  the  scanning  electron 
microscope. 

Because  hydrofluoric  acid  is  extremely  dangerous,  the  fluoridization  process  should  always  be  carried  out  in  a 
properly  designed  fume  cupboard  with  an  efficient  extraction  system,  by  an  operator  wearing  protective 
clothing,  rubber  gloves,  and  a face-mask.  The  reaction  between  the  sample  and  the  acid  may  be  very  vigorous, 
and  large  amounts  of  carbon  dioxide  may  be  generated  rapidly.  It  is  important,  therefore,  to  treat  the  sample 
in  an  adequately  large  polythene  vessel  to  prevent  froth  from  forming  and  spilling  out.  We  fluoridize 
approximately  10  g of  bulk  sample  in  an  80  mm-diameter  400  ml  polythene  beaker. 


COMMENTS  ON  THE  SPECIMENS  ILLUSTRATED 

The  four  ossicles  illustrated  in  Plates  94  and  95  were  obtained  from  a bulk  sample  of  the  soft  clay  shale 
above  the  Charlestown  Main  Limestone,  collected  near  the  bathing  pool,  St.  Monance,  Fife, 
Scotland  (National  Grid  Reference  NO  536  020).  The  shale  is  of  Lower  Carboniferous  (Brigantian) 
age  and  has  been  correlated  with  the  Neilson  Shell  Band  (George  et  al.  1976,  fig.  14,  p.  53).  The 
sample  was  partly  disaggregated  by  being  soaked  in  paraffin,  and  then  boiled  in  water  with  soda 
ash.  Small  amounts  of  the  disaggregated  material  were  reacted  with  48%  HF  for  1 hour.  The 
fluoridized  ossicles  were  mounted  on  stubs  and  coated  with  carbon  and  a gold  palladium 
mixture,  and  were  examined  using  an  ETEC  Autoscan,  Model  H-l,  scanning  electron  microscope. 
The  diameters  of  stereom  pores  were  measured  on  enlarged  scanning  electron  micrographs  and  the 
surface  porosity  by  point  counting  along  two  mutually  perpendicular  axes  as  suggested  by  Macurda 
and  Meyer  (1975,  p.  2).  The  terminology  used  is  from  Ubaghs  (1978,  T.  58  et  seq.).  The  illustrated 
specimens  and  other  representative  material  are  reposited  in  the  palaeontological  collections  of 
Trinity  College,  Dublin  (catalogue  numbers  prefixed  TCD). 

Pentagonal  columnals  (TCD  19861-3)  (PI.  94,  figs.  1,  3) 

Columnals  of  this  kind  are  moderately  abundant  in  the  sample.  The  longest  pluricolumnal  found 
consists  of  a nodal  between  two  pairs  of  internodals.  The  nodal  is  cirrus-bearing  and  approximately 


SEVASTOPULO  AND  KEEGAN:  STEREOM  STRUCTURE  OF  FOSSIL  CRINOIDS  751 

0-8  times  as  long  as  wide;  the  internodals  are  of  two  orders  with  length  to  width  ratios  of  0-4  and  0-6. 
The  sides  of  the  columnals  are  straight,  or  have  a ridge  or  swelling  around  the  equator,  a feature 
particularly  well  developed  on  the  nodals.  Each  nodal  has  one  or  two  cirral  sockets  positioned 
between  the  equator  and  the  joint  surface.  The  sockets  are  comparable  in  some  respects  to  cirral 
facets  of  Mesozoic  crinoids  illustrated  by  Ubaghs  (1978,  T.  85,  fig.  61).  They  are  gently  concave  and 
slope  towards  the  joint  face.  The  lumen  of  the  axial  canal  is  a vertical  slit.  The  half  of  the  socket  closest 
to  the  equator  of  the  columnal  is  furnished  with  short  culmina;  the  half  closest  to  the  joint  face  is 
smooth.  The  sides  of  the  columnal  are  formed  of  dense  stereom  with  a surface  ornament  of  slightly 
raised  granules  approximately  1 5 /xm  in  diameter. 

In  facetal  view  (PI.  94,  fig.  1)  the  following  regions  of  the  articulum  can  be  differentiated: 

1 . The  lumen,  approximately  20-25%  of  the  width  of  the  articulum,  which  appears  faintly  five-  or 
ten-lobed  in  well-preserved  specimens. 

2.  An  adaxially  sloping  concave  area  surrounding  the  lumen  (the  floor  of  the  spatium), 
approximately  10-12%  of  the  width  of  the  articulum.  The  degree  to  which  this  region  is  depressed  is 
variable;  it  is  very  shallow  in  the  specimen  illustrated.  It  is  floored  by  open  stereom  (round  to  ovoid 
pores,  with  diameters  from  6 to  14  /xm,  mostly  about  12  /xm)  which  in  broken  specimens  can  be  seen  to 
form  a thin  layer  overlying  denser  paraxial  galleried  stereom  like  that  flooring  the  areola. 

3.  In  some  specimens  (but  not  the  figured  example)  the  outer  margin  of  the  floor  of  the  spatium  is 
raised  to  form  a narrow  perilumen  constructed  of  denser  stereom. 

4.  The  areola  (approximately  10-15%  of  the  width  of  the  articulum)  which  is  flat  and  floored  by 
paraxial  galleried  stereom  (pore  diameter  6-9  /xm;  surface  porosity  approximately  44%).  Most  pores 
are  subrounded  and  bounded  by  four  trabeculae  and  many  are  arranged  in  long  slightly  arcuate  rows. 

5.  The  crenularium  (approximately  10-15%  of  the  width  of  the  articulum)  consisting  of  steep- 
sided culmina  and  crenellae  (PI.  94,  fig.  3).  The  top  of  the  culmina  and  base  of  the  crenellae  are 
approximately  equidistant  from  the  level  of  the  areola.  The  surfaces  of  the  culmina  are  dense  with 
conspicuously  thickened  trabecular  intersections  (pore  diameters  are  2-5-5-0  /xm;  surface  porosity 
30%  or  less),  but  are  underlain  by  paraxial  galleried  stereom.  The  crenellae  are  mostly  floored  by 
galleried  stereom  similar  to  that  of  the  areola  (pore  diameters  7-10  /xm),  but  in  some  the  stereom  is 
much  more  open  and  labyrinthic. 

A conspicuous  feature  of  the  articulum  is  the  set  of  large  tunnel-like  pores  (up  to  35  /xm  in  diameter) 
which  in  several  specimens  can  be  seen  to  completely  penetrate  the  columnal.  They  are  crudely 
arranged  in  ten  lines  and  extend  to  the  outer  part  of  the  areola. 

In  most  respects  the  microstructure  of  these  Carboniferous  columnals  is  comparable  with  that  of 
Recent  and  Mesozoic  columnals  described  by  Macurda  and  Meyer  (1975,  1976)  and  Roux  (1971). 
The  galleried  stereom  of  the  areola  probably  housed  ligament  fibres.  The  denser  stereom  of  the 
perilumen  and  of  the  crenularium  served  as  bearing  surfaces.  The  large  pores  penetrating  the 
columnals  may  have  contained  nerves,  as  suggested  by  Macurda  and  Meyer  (1975,  p.  3)  for  similar 
pores  in  the  columnals  of  the  Recent  species  Isocrinus  blakei.  The  pore  diameters  of  the  columnals 
described  here  are  consistently  smaller  than  those  reported  for  most  Recent  and  Mesozoic  forms. 

The  taxonomic  affinity  of  the  specimens  is  not  known.  They  almost  certainly  belonged  to  a cladid 
inadunate,  possibly  an  ampelocrinid  in  view  of  the  pentagonal  stem  and  cirrus-bearing  nodals. 

Elliptical  columnal  of  Platycrinites  (TCD  19864-6)  (PI.  94,  figs.  2,  4) 

Columnals  with  elliptical  articular  surfaces  are  moderately  common  in  the  sample.  They  vary 
considerably  in  shape.  The  majority,  mainly  smaller  specimens,  are  longer  than  wide  and  have  a 
distinct  equatorial  waist.  Most  of  them  bear  scattered  nodes  or  blunt  spines.  A few  specimens  are 
wider  than  long,  and  some  of  these,  possibly  nodals,  have  conspicuous  equatorial  spine-bearing 
flanges.  The  articular  surfaces  are  also  variable  although  a basic  pattern  can  be  observed  in  all  of 
them:  a raised  fulcral  region  along  the  major  axis  of  the  face  separates  two  gently  concave  fields.  The 
lumen  is  small  and  elliptical  and  is  surrounded  by  open  paraxial  galleried  stereom  (pore  diameters  up 
to  1 3 /xm;  porosity  approximately  37%).  The  central  parts  of  the  bifascial  fields  are  floored  by  paraxial 


752 


PALAEONTOLOGY,  VOLUME  23 


galleried  stereom  with  pore  diameters  typically  6-10  ^ m and  porosity  approximately  32%.  The 
peripheries  of  the  faces  are  slightly  raised  above  the  bifascial  fields  and  are  formed  of  denser  stereom 
(pore  diameter  3-5  /xm;  porosity  less  than  30%).  The  long  axis  of  the  articular  surface  is  occupied  by  a 
fulcral  region  which  in  many  specimens  consists  of  a broad  slightly  raised  ridge  of  dense  stereom 
(pore  diameter  typically  4 ^m;  surface  porosity  approximately  20%).  In  some  specimens  the  surface  of 
the  fulcral  region  is  crossed  by  low,  dense,  vermiform  ridges.  At  each  end  of  the  major  axis  of  the 
articular  surface  are  raised  culmina,  generally  three  in  number,  which  rise  above  the  level  of  the 
fulcral  region  (PI.  94,  fig.  4).  They  interlock  with  crenellae  of  adjacent  columnals.  The  culmina  are 
formed  of  dense  stereom  (pore  diameters  typically  less  than  4 /xm;  porosity  less  than  20%)  and  the 
crenellae  are  floored  by  galleried  stereom  (pore  diameter  typically  9 p.m).  The  major  axes  of  opposing 
faces  of  many  of  the  columnals  are  set  at  90°  to  each  other. 

The  ossicles  are  easily  identified  as  belonging  to  Platycrinites  but  their  specific  identity  is  not 
known.  In  many  respects  their  structure  is  comparable  with  that  of  the  columnals  of  the  Recent 
millericrinid  Democrinus  (Macurda  and  Meyer  1975,  pp.  4,  5)  which  also  has  synarthrial  articulation. 
In  the  Scottish  Platycrinites,  however,  the  fulcral  ridge  is  much  less  dense  than  in  Democrinus,  and 
the  elaborate  keying  mechanisms  of  that  genus  are  not  developed.  Instead,  limited  symplectial 
articulation  occurred  at  both  ends  of  the  fulcral  ridge. 

Inadunate  brachial  (TCD  19867-9)  (PI.  95,  figs.  1,  3) 

This  kind  of  brachial  is  the  most  common  in  the  sample.  All  examples  that  have  been  found  are 
cuneate,  pinnule-bearing,  higher  than  long,  and  most  have  nodes  or  blunt  spines  on  the  aboral 
surface,  particularly  along  the  distal  margins.  All  the  brachials  were  joined  by  oblique  muscular 
articulations;  the  fulcral  ridges  on  the  two  faces  of  a brachial  may  diverge  by  as  much  as  60°.  The 
following  regions  may  be  differentiated  on  the  articular  surfaces  (PI.  95,  fig.  1): 

1.  The  fulcral  ridge,  which  is  narrow  at  its  mid-point  and  widens  slightly  at  both  ends  to 
approximately  75  ^m  in  typical  specimens.  The  ridge  is  constructed  of  dense  stereom  (pore  diameters 
typically  4 ^ m or  less;  porosity  approximately  20%). 

2.  A slightly  depressed  area  less  than  30  /xm  deep  bounded  by  the  fulcral  ridge  and  the  aboral 
margin.  By  analogy  with  Recent  crinoids,  this  area  in  Palaeozoic  inadunates  has  been  identified  as  the 
aboral  ligament  fossa  which  housed  the  extensor  ligament  bundles.  It  is  floored  by  galleried  stereom 
(pore  diameters  typically  7 /x m;  porosity  approximately  35%).  The  pores  are  subrounded  and 
arranged  in  a crude  rectilinear  pattern.  In  most  specimens  (but  not  the  figured  example)  a distinct 
small  deeper  ligament  pit  occurs  just  aborally  of  the  mid-point  of  the  fulcral  ridge. 

3.  Two  wide  subequal  depressions,  typically  less  than  30  /x m deep,  adoral  of  the  fulcral  ridge  and 
on  either  side  of  its  mid-point,  which  have  been  identified  as  interarticular  ligament  fossae.  They 
are  floored  by  galleried  stereom  in  which  the  trabeculae  and  pores  are  conspicuously  wider  than 
elsewhere  on  the  articular  surface.  Pore  diameters  generally  range  from  10  to  15  /xm;  the  porosity  is 
approximately  40%. 


EXPLANATION  OF  PLATE  94 

Figs.  1,  3.  Fluoridized  pentagonal  columnal  (TCD  19861),  from  the  shale  above  the  Charlestown  Main 
Limestone,  St.  Monance,  Fife  (Lower  Carboniferous;  Brigantian  age).  1,  slightly  oblique  view  of  the  articular 
surface,  x 45.  3,  stereopair  of  the  crenularium  and  outer  part  of  the  areola,  located  at  about  7 o’clock  on 
fig.  1,  x 230. 

Figs.  2,  4.  Fluoridized  Platycrinites  columnal  (TCD  19864),  from  the  shale  above  the  Charlestown  Main 
Limestone,  St.  Monance,  Fife  (Lower  Carboniferous;  Brigantian  age).  2,  oblique  view  of  columnal,  x 38.  4, 
stereopair  of  part  of  the  fulcral  ridge  and  culmina,  from  the  left  side  of  fig.  2,  x 1 50. 


PLATE  94 


imm 


Mmm 


l wmM 


;*®CS 


SEVASTOPULO  and  KEEGAN,  Stereom  structure  of  fossil  crinoids 


754 


PALAEONTOLOGY,  VOLUME  23 


4.  A slightly  raised  area  extending  from  the  adoral  groove  to  the  mid-point  of  the  fulcral  ridge  and 
separating  the  interarticular  ligament  fossae.  This  area  bears  a very  weak  medial  groove  which  ends 
short  of  the  fulcral  ridge.  The  stereom  of  the  raised  area  is  galleried  (pore  diameter  6-8  jxm)  along  the 
margins  and  more  open  (pore  diameter  10-15  pm)  and  less  regular  along  the  median  groove. 

5.  Well-marked,  unequal,  ‘rabbit-ear’-shaped  depressions  on  either  side  of  the  ambulacral  groove, 
which  have  been  interpreted  as  flexor  muscle  scars.  They  are  floored  by  distinctive  dense  labyrinthic 
stereom  (pore  diameters  mostly  less  than  4 pm;  porosity  approximately  20%).  The  surfaces  of  the 
fossae  are  formed  by  blunt-ended  trabecular  rods  projecting  upwards  (PI.  95,  fig.  3). 

All  well-preserved  specimens  can  be  seen  to  have  three  pores  20-30  pm  in  diameter  adoral  of  the 
fulcral  ridge  on  both  articular  surfaces.  Two  of  them  lie  along  a line  normal  to  the  bisectrix  of  the 
angle  of  the  adoral  groove;  the  third  is  between  the  other  two,  closer  to  the  fulcral  ridge. 

Many  of  the  features  observed  are  similar  to  those  reported  by  Lane  and  Macurda  (1975)  for  the 
Pennsylvanian  cladid  inadunate  Aesiocrinus  magnificus.  The  upward  projecting  trabecular  rods  of 
the  ‘rabbit  ear’  fossae  were  probably  sheathed  with  a thin  connective  tissue  layer  to  which  the  muscle 
fibres  were  attached,  as  illustrated  for  the  Recent  crinoid  Annacrinus  by  Roux  (19746,  pi.  1,  figs.  6-7). 
An  unusual  feature  of  the  Scottish  brachials  is  the  presence  of  the  three  canals  interpreted  here  as 
aboral  nerve  canals.  In  Recent  crinoids  there  is  only  one  canal  in  the  brachials;  Lane  and  Macurda 
(1975)  showed  that  in  Aesiocrinus  a ‘double-barelled’  nerve  canal  was  present.  We  have  found  the 
‘double-barelled’  arrangement  in  a number  of  different  inadunate  brachials,  but  the  triple  canal  has 
only  been  found  in  the  ossicles  described  here.  We  are  unable  to  identify  the  brachials.  They  clearly 
were  from  a cladid  inadunate.  We  have  recovered  axillary  brachials,  which  show  that  the  rays  were 
branched  and  that  the  first  dichotomy  was  above  the  first  primibrachial. 

Flexible  brachial  { TCD  19870-3)  (PI.  95,  figs.  2,  4) 

Brachials  of  this  kind  are  moderately  abundant  in  the  sample,  but  there  is  considerable  variation  in 
the  ratio  of  width  to  height;  possibly  more  than  one  crinoid  species  is  represented.  The  proximal 
articular  surfaces  are  extended  aborally  into  patelloid  processes  and  the  distal  surfaces  each  have  a 
fossa  into  which  the  process  fits.  In  the  specimen  illustrated  (PI.  95,  fig.  2),  the  lateral  margins  of  the 
articular  surface  are  crenulate  with  steep-sided  culmina  approximately  100  pm  high.  There  is  no 
fulcral  ridge,  but  a fulcral  ridge  is  present  on  some  larger  specimens.  The  stereom  of  the  articular 
surface  occurs  in  three  different  arrays.  Over  most  of  the  surface,  excluding  the  area  around  the 
patelloid  process  and  a narrow  median  area  extending  aborally  from  the  adoral  groove,  the  pores  are 
quadrangular  to  round,  the  diameters  of  13-20  pm,  and  the  porosity  is  approximately  45%.  The 
trabeculae  on  either  side  of  the  aboral/adoral  axis  of  the  brachial  are  oriented  at  similar  angles  to  the 
median  line  and  produce  a markedly  rectilinear  pore  pattern.  The  stereom  pores  visible  in  side  view 
are  approximately  the  same  dimensions  as  on  the  articular  surfaces,  so  that  the  trabeculae  form 
a regular  cubic  framework.  In  the  median  region,  aboral  of  the  adoral  groove,  the  pores  are  slightly 
reduced  in  size  and  the  regular  arrangement  of  the  pores  is  lost.  On  the  aboral  part  of  the  articular 
surface  around  the  patelloid  process,  the  stereom  is  much  denser  (pore  diameter  5-8  pm;  porosity  less 
than  30%)  and  less  regularly  arranged. 


EXPLANATION  OF  PLATE  95 

Figs.  1,  3.  Fluoridized  inadunate  crinoid  brachial  (TCD  19867),  from  the  shale  above  the  Charlestown  Main 
Limestone,  St.  Monance,  Fife  (Lower  Carboniferous;  Brigantian  age).  1,  slightly  oblique  view  of  proximal 
articular  surface,  with  pinnule  facet  to  the  left,  x 48.  3,  stereopair  of  the  left  side  ‘rabbit  ear’  fossa  and 
adjoining  interarticular  ligament  fossa,  x 300. 

Figs.  2, 4.  Fluoridized  flexible  brachial  (TCD  19870),  from  the  shale  above  the  Charlestown  Main  Limestone,  St. 
Monance,  Fife  (Lower  Carboniferous;  Brigantian  age).  2,  view  of  the  proximal  articular  surface,  x48.  4, 
stereopair  of  culmina,  located  near  the  middle  of  the  left  margin  in  fig.  1,  x 180. 


PLATE  95 


SEVASTOPULO  and  KEEGAN,  Stereom  structure  of  fossil  crinoids 


756 


PALAEONTOLOGY,  VOLUME  23 


These  flexible  brachials  cannot  be  more  closely  identified;  all  the  Carboniferous  flexible  ossicles 
encountered  in  this  study  have  had  remarkably  similar  microstructure. 

Most  authors  (for  instance.  Van  Sant  and  Lane  1964,  p.  51)  have  suggested  that  flexible  crinoids 
had  only  ligamentary  articulations.  Whether  ligament  fibres  penetrated  all  the  ‘cubic’  structured 
stereom  or  were  restricted  to  certain  areas  is  not  certain,  but  the  former  arrangement  seems  more 
likely. 

Acknowledgements.  This  study  was  completed  when  one  of  us  (G.  D.  S.)  was  visiting  the  Department  of  Geology, 
Indiana  University,  Bloomington,  Indiana,  U.S.A.  We  thank  Haydn  Murray,  Chairman,  for  providing 
facilities,  and  members  of  the  department  for  discussion  and  other  help.  N.  Gary  Lane  kindly  read  the 
manuscript  which  was  typed  by  Sandy  K.  Douthitt.  Herschel  Lentz,  Department  of  Biology,  Indiana  University, 
helped  with  the  scanning  electron  microscopy.  We  thank  R.  B.  Wilson,  Institute  of  Geological  Sciences, 
Edinburgh,  for  information  regarding  the  St.  Monance  locality.  The  study  arose  out  of  a micropalaeontological 
project  supported  by  the  Irish  National  Board  for  Science  and  Technology. 

REFERENCES 

cookson,  I.  c.  and  singleton,  o.  p.  1954.  The  preparation  of  translucent  fossils  by  treatment  with  hydrofluoric 
acid.  Geol.  Soc.  of  Australia  News  Bulletin,  2,  1 -2. 

GEORGE,  T.  N.,  JOHNSON,  G.  A.  L.,  MITCHELL,  M.,  PRENTICE,  J.  E.,  RAMSBOTTOM,  W.  H.  C.,  SEVASTOPULO,  G.  D.  and 

wilson,  R.  b.  1976.  A correlation  of  Dinantian  rocks  in  the  British  Isles.  Geol.  Soc.  Lond.  Special  Report  No.  7, 
87  pp. 

Grayson,  J.  F.  1956.  The  conversion  of  calcite  to  fluorite.  Micropaleontology,  2,  71-78. 

lane,  n.  G.  and  macurda,  D.  B.,  Jun.  1975.  New  evidence  for  muscular  articulations  in  Paleozoic  crinoids. 
Paleobiology,  1,  59-62. 

— and  sevastopulo,  G.  D.  (in  press).  Functional  morphology  of  a microcrinoid:  Kallimorphocrinus  punctatus 
n.  sp.  J.  Paleont. 

lapham,  K.  e.,  ausich,  w.  i.  and  lane,  N.  G.  1976.  A technique  for  developing  the  stereom  of  fossil  crinoid 
ossicles.  Ibid.  50,  245-248. 

macurda,  d.  b.,  Jun.  and  meyer,  d.  l.  1975.  The  microstructure  of  the  crinoid  endoskeleton.  Paleont.  Contrib. 
Univ.  Kansas,  74,  1-22,  pis.  1-30. 

— 1976.  The  morphology  and  life  habits  of  the  abyssal  crinoid  Bathycrinus  aldrichianus  Wyville  Thomson  and 
its  palaeontological  implications.  J.  Paleont.  50,  647-667,  pis.  1-5. 

roux,  m.  1970.  Introduction  a l’etude  des  microstructures  des  tiges  de  crino'ides.  Geobios.  3,  79-98,  pis.  14-16. 

— 1971.  Recherches  sur  la  microstructure  des  pedoncules  de  crinoides  post-Paleozoiques.  Univ.  Paris,  Fac. 
Sci.  Orsay,  Trav.  Lab.  Paleontol.,  83  pp.,  4 pi. 

— 1974a.  Les  principaux  modes  d’articulation  des  ossicules  du  squelette  des  Crinoides  pedoncules  actuel. 
Observation  microstructurales  et  consequences  pour  f interpretation  des  fossiles.  Acad.  Sci.  Paris,  Comptes 
Rendus,  278(D),  2015-2018. 

— 19746.  Observations  au  microscope  electronique  a balayage  de  quelque  articulations  entre  les  ossicules  du 
squelette  des  Crinoides  pedoncules  actuels  (Bathycrinidae  et  Isocrinina).  Univ.  Paris,  Fac.  Sci.  Orsay,  Trav. 
Lab.  Paleontol.,  9 pp.,  4 pi. 

— 1975.  Microstructural  analysis  of  the  crinoid  stem.  Paleont.  Contrib.  Univ.  Kansas,  75,  1-7,  pis.  1-2. 
sohn,  i.  G.  1956.  The  transformation  of  opaque  calcium  carbonate  to  translucent  calcium  fluoride  in  fossil 

Ostracoda.  /.  Paleont.  30,  113-114,  pi.  25. 

sprinkle,  j.  and  gutschick,  r.  c.  1967.  Costatoblastus,  a channel  fill  blastoid  from  the  Sappington  Formation  of 
Montana.  Ibid.  41,  385-402,  pi.  45. 

ubaghs,  G.  1978.  Skeletal  morphology  of  fossil  crinoids.  In  moore,  r.  c.  and  teichert,  c.  (eds.).  Treatise  on 
invertebrate  paleontology.  New  York  and  Lawrence,  Geol.  Soc.  Am.,  pt.  T,  Echinodermata  2,  1,  T58-T216. 
upshaw,  c.  F.,  todd,  R.  G.  and  allen,  b.  d.  1957.  Fluoridization  of  microfossils. ./.  Paleont.  31,  793-795,  pi.  100. 
van  sant,  J.  F.  and  lane,  N.  G.  1964.  Crawfordsville  (Indiana)  crinoid  studies.  Univ.  Kansas,  Paleont.  Contrib., 
Echinodermata  Art.  7,  pp.  1-136,  pis.  1-8. 

wetzel,  w.  1921 . Darstellung  von  Flusspat  bei  Zimmertemperatur.  Centralbl.f.  Min.,  Geol.  u.  Palaont.  444-447. 
zingula,  R.  p.  1968.  A new  breakthrough  in  sample  washing.  J.  Paleont.  42,  1092. 


Typescript  received  12  September  1979 


G.  D.  SEVASTOPULO,  J.  B.  KEEGAN 
Department  of  Geology 
Trinity  College,  Dublin  2 


THE  VALUE  OF  OUTLINE  PROCESSING  IN  THE 
BIOMETRY  AND  SYSTEMATICS  OF  FOSSILS 

by  G.  H.  SCOTT 


Abstract.  Widespread  use  of  gross  dimensions  and  similar  point-to-point  measurements  in  biometric  studies  of 
fossils  is  probably  due  more  to  instrumental  limitations  and  the  influence  of  preceding  studies  than  to  theoretical 
considerations.  Are  such  data  suitable  for  classificatory  studies  which  are  heavily  dependent  on  visual 
assessment  of  morphology?  Theory  suggests  that  the  outlines  of  objects  are  particularly  significant  in  visual 
recognition  because  of  their  high  information  content.  They  provide  a parsimonious  description  of  form. 
Biometry  can  best  supplement  qualitative  visual  processes  in  taxonomic  studies  by  treating  outline  data  in  ways 
that  replace  the  information  lost  due  to  the  short-term,  degradable  nature  of  visual  data  stored  in  the  human 
memory.  Variation  in  the  axial  outlines  of  the  foraminifer  Globorotalia  puncticulata  (Deshayes)  is  examined  as 
an  example. 

Data  collection  is  fundamental  to  biometry.  Nevertheless,  textbooks  concentrate  on  techniques  of 
data  reduction  and  analysis,  and  offer  little  guidance  about  the  collection  of  data.  Such  limited 
reference  is  understandable.  Organisms  are  exceedingly  diverse  in  form  and  organization.  Guidelines 
for  the  collection  of  quantitative  data  can  be  cited  (e.g.  Simpson,  Roe,  and  Lewontin  1960)  but 
concepts  such  as  ‘character’  and  ‘variable’  are  so  context-dependent  that  most  writers  seem  to 
concede,  at  least  implicitly,  that  their  selection  in  biometric  studies  should  be  left  to  the  discretion  of 
the  student.  While  the  literature  indicates  that  there  is  considerable  accord  among  researchers  on 
protozoans  to  vertebrates  on  the  types  of  data  to  be  collected,  this  does  not  necessarily  signify 
adherence  to  a common  rationale  of  data  collection.  Precedents  and  instrumental  constraints  exert 
powerful  influences  on  the  data  collected  in  a project.  Here  I consider  the  role  of  biometry  in 
classificatory  studies  (broadly,  recognition  of  taxa  and  allocation  of  specimens)  in  the  light  of  theory 
on  the  mechanisms  of  visual  perception.  It  is  advocated  that  biometry  should  supplement  these 
mechanisms  by  processing  comparable  data  so  that  there  is  a parallel  between  qualitative  and 
quantitative  treatments  of  specimens.  In  this  way  biometry  can  contribute  to  resolving  the  problems 
of  the  systematist  that  arise  from  deficiencies  in  visual  recognition. 

TYPICAL  PRACTICE 

While  it  is  not  claimed  that  the  measurements  illustrated  in  text-fig.  1 portray  all  aspects  of  modern 
practice  in  variate  selection,  they  are  sufficiently  representative  to  indicate  that  biometric  studies 
primarily  use  data  on  the  gross  dimensions  of  structures.  Point-to-point  measurements  of  maximum 
dimensions  of  skeletal  parts  form  the  vast  majority  of  the  data  reported  in  the  literature,  and  the 
example  of  the  measurement  of  the  length  of  a curve  (text-fig.  lc)  is  unusual. 

Instrumentation,  operational  convenience,  and  the  precedents  set  by  previous  studies  account  for 
the  preference  for  gross  dimensions.  The  first  two,  in  conjunction,  are  fundamental.  Operationally, 
gross  dimensions  offer  considerable  advantages  in  variate  selection.  Much  of  the  form  of  skeletal 
structures  consists  of  smooth,  continuously  curved  surfaces.  In  such  regions  well-defined,  relocatable 
loci  for  measurement  may  be  few,  and  the  obvious  ‘landmarks’  for  the  biometrician  are  the 
extremities  of  the  structure.  Usually  these  are  homologous  within  the  population  sampled.  The 
simple  scales  and  calipers  which  are  the  stock  in  trade  of  the  palaeontologist  are  well  suited  to 
measurements  of  gross  dimensions,  whereas  they  are  unsuited  to  determining  the  lengths  of  vectors  or 
curves,  for  example.  Indeed,  the  widespread  use  of  gross  dimensions  and  of  measurements  between 

IPalaeontology,  Vol.  23,  Part  4,  1980,  pp.  757-768.] 


758 


PALAEONTOLOGY,  VOLUME  23 


well-defined  ‘landmarks’  in  biometry  is  probably  due  as  much  to  the  limitations  of  instruments  as  to 
their  value  on  purely  biological  grounds  in  morphological  description  and  analysis. 

Precedent  is  an  ancillary  influence  that  tends  to  stabilize  the  set  of  characters  measured  and 
perhaps  inhibits  fresh  consideration  of  what  should  be  measured.  Moreover,  pioneering  works  that 
use  point-to-point  measurements  have  an  advantage  in  the  selection  of  precedents  because  of  the 
general  availability  of  comparable  measurement  devices.  The  measurements  (partly  shown  in  text-fig. 
1a)  on  trilobites  made  by  Shaw  (1957)  are  a good  example  of  the  influence  on  later  workers  (Temple 
1975)  of  a pioneering  study. 


A BASIS  FOR  BIOMETRY 

There  is  very  little  evidence  in  the  literature  that  theoretical  considerations  have  influenced  the  choice 
of  characters  for  measurement.  In  an  introduction  to  a major  biometric  study  of  Ostracoda,  Reyment 
(1963)  asserted  that  statistical  analysis  would  provide  a comprehensive  representation  of  variation 
but  made  no  comment  on  the  adequacy  of  the  measurements  (carapace  length,  height,  and  breadth) 
that  formed  the  great  majority  of  his  data.  It  is  conjectural  whether  the  claim  by  Hallam  and  Gould 
(1975,  p.  517)  that  their  nine  measurements  on  the  left  valve  of  Gryphaea  are  ‘adequate  to  express 
overall  features  of  valve  shape  and  the  character  of  the  sulcus’  can  be  substantiated.  Most  workers 
(myself  included)  can  be  easily  pilloried  on  the  grounds  of  ad  hoc  selection  of  data  without 
justification.  A relevant -example  is  Melville’s  (1978)  critique  of  a biometric  study  of  leaf  shape  in 
Ulmus.  Which  features  should  be  selected  for  measurement? 

The  choice  of  measurements  and  their  analysis  should  relate  to  the  aims  and  methods  of  the 
investigation.  This  is  self-evident  in  an  application  of  biometry  to  a study  of  functional  morphology, 
for  example,  where  mechanical  hypotheses  are  presented  for  testing.  But  it  is  a useful  point  of 
departure  when  considering  the  role  of  biometry  in  the  generally  less-structured  tasks  of  classifica- 
tion. Here  the  primary  activities  concern  the  establishment  of  classes  and  the  allocation  of  specimens. 
The  principal  problems  concern  the  estimation  of  intra-group  variation  and  inter-group  separation 
or  distance.  Where  do  class  limits  fall?  Modern  evolutionary  theory  and  research  provide  a cogent 
account  of  the  mechanisms  of  variation.  The  systematist,  however,  is  presented  with  the  end  products 
of  various  genetic,  phenotypic,  ontogenetic,  and  diagenetic  processes.  In  a particular  instance  there 
may  be  very  strong  reasons,  a priori , to  suppose  that  the  specimens  under  systematic  scrutiny  are 
samples  from  discrete  populations.  The  problem  is  that  of  recognition. 

Although  data  on  distribution  and  ecology  are  significant,  the  primary  information  in  the 
systematics  of  fossils  is  morphological,  obtained  by  qualitative  visual  examination.  The  immense 
production  of  illustrations  of  fossils  over  the  last  two  centuries  attests  the  fundamental  importance  of 
visual  representation  in  systematics.  Certainly,  the  initial  phase  of  simple  qualitative  visual 
assessment  is  followed  by  analysis,  sometimes  using  quantitative  data,  that  leads  to  diagnoses  of  taxa. 
But  the  latter  is  a conscious  refinement  of  the  initial  phase.  The  brain  is  an  immensely  fast  and 
powerful  processor  of  visual  imagery.  Visual  data  are  rapidly  assembled,  images  reconstructed  and 
interpreted.  Messages  about  the  identity  of  specimens  are  produced  almost  involuntarily  and  are  the 
basis  of  classificatory  work.  The  process  is  that  used  in  other  visual  recognition  tasks  in  day-to-day 
experience,  although  a higher  standard  of  recognition  and  discrimination  is  desirable.  Form 
variation  in  biological  materials  is  often  complex,  with  major  ontogenetic  and  environmental  sources 
to  be  allowed  for  in  taxonomic  recognition. 

What  is  the  role  of  biometry  in  such  studies?  Should  it  supplant  or  supplement  qualitative 
perception?  If  only  for  reasons  of  instrumentation,  the  present  role  must  be  supplementary.  In  many 
aspects  the  human  visual  system  is  more  advanced  than  any  similar  device.  It  is  in  inter-image 
discrimination  that  the  human  system  is  least  effective,  especially  when  sample  sizes  are  large, 
variation  multidimensional,  and  groups  ill  defined.  Objects  are  scanned  and  features  of  others 
recalled  in  attempts  to  reach  classificatory  decisions.  Here  the  static,  long-term  memories  of  digital 
devices  seem  to  have  marked  advantages  over  the  human  system.  Re-recording  of  image  information 
to  refresh  the  memory  is  made  unnecessary.  Once  stored,  it  remains  available  for  recall  and 


SCOTT:  OUTLINE  PROCESSING 


759 


text-fig.  1.  Measurements  of  structures  commonly  preserved  as  fossils.  Variate 
identifications  and  scales  are  omitted  in  the  adaptations,  a,  non-agnostidean 
trilobite  cephalon  after  Shaw  (1957,  text-fig.  11).  b,  gastropod,  Athleta petrosa 
(Conrad),  after  Fisher,  Rodda,  and  Dietrich  (1964,  text-fig.  1).  c,  bivalve, 
Gryphaea,  after  Hallam  and  Gould  (1975,  fig.  1).  D,  pterosaur  skull, 
Pterodactylus,  after  Mateer  (1976,  fig.  1).  e,  conodont,  after  Sergeyeva  et  al. 
(1975,  fig.  5).  F,  acritarch,  after  Sellberg  and  Kjellstrom  (1975,  fig.  1).  G, 
brachiopod,  Linnarssonella  girtyi  Walcott,  after  Rowell  (1966,  table  4).  H, 
ostracod,  Bairdia  victrix  Brady,  after  Cadot  and  Kaesler  (1973,  fig.  2).  i, 
foraminifer,  Globorotalia  miozea  miozea  Finlay,  after  Scott  (1972,  text-fig.  2).  j, 
ammonite,  Vascoceras,  after  Berthou,  Brower,  and  Reyment  (1975,  fig.  c).  K, 
molar  teeth  of  condylarth  mammal,  after  Olson  and  Miller  (1958,  fig.  61).  L, 
amphibian  skull,  Trimerorhachis,  after  Olson  (1953,  fig.  1). 


760 


PALAEONTOLOGY,  VOLUME  23 


reprocessing  without  degradation.  If  biometry  is  to  supplement  the  ‘weak’  points  of  visual 
perception,  it  follows  that  it  should  process  the  same  sort  of  data.  The  problem  with  ad  hoc  characters 
is  that  they  may  record  aspects  of  the  object  that  are  insignificant  in  visual  processing.  How  does  the 
human  system  function? 

Visual  perception.  Once  the  preserve  of  the  psychologist,  the  mechanics  of  visual  perception  have 
become  an  interdisciplinary  subject  because  of  their  relevance  in  automatic  pattern  recognition  and 
allied  studies.  A comprehensive  survey  is  not  attempted,  but  there  is  general  agreement  about  the 
significance  of  the  outline  in  object  recognition.  Gestalt  psychologists  (e.g.  KofFka  1935)  con- 
centrated on  those  properties  of  figures  that  facilitated  their  recognition  or  isolation  from 
background  data.  One  of  their  laws  of  organization  drew  attention  to  the  importance  of  closure. 
Closed  figures  tend  to  be  perceived  as  units  more  readily  than  unclosed.  From  quite  different 
premises,  information  theorists  showed  that  much  visual  data  is  redundant  in  recognition  processes 
because  of  high  correlation  among  the  data  received  by  adjacent  visual  receptors.  Attneave  (1954) 
gave  a simple,  convincing,  example  of  this  and  suggested  that  early  visual  processing  filters  out  much 
redundant  information,  leaving  a reduced,  more  economic,  description  of  the  data.  Redundancy  is 
high  in  regions  of  an  object  that  are  homogeneous  in  some  visual  property  (e.g.  colour,  texture, 
curvature)  and  low  in  regions  where  such  properties  change  rapidly.  The  margins  of  an  object  are 
regions  where  redundancy  is  particularly  low,  although  zones  of  uniform  slope  or  curvature  along  the 
margin  have  higher  redundancy  than  those  in  which  there  are  rapid  changes  in  direction  or  slope. 
Attneave  showed  that  an  object  can  be  recognized  readily  from  a simplified  sketch  consisting  of  the 
points  of  maximum  curvature  of  the  outline  linked  by  straight  lines.  Such  a result  is  an  explanation  of 
the  verisimilitude  achieved  so  effortlessly  by  the  competent  cartoonist  or  street  artist.  But  it  is  also 
highly  suggestive  to  the  biometrician.  Marr  (1976)  suggested  that  a major  element  in  early  visual 
processing  is  the  construction  of  a ‘primal  sketch’  from  grey-level  changes  in  the  receptor  data  array. 
Intensity  changes  are  isolated  and  used  to  construct  a description  of  the  array.  Edges  are  major 
elements  in  the  description. 

Commentary.  The  review  indicates  the  prime  importance  of  outline  data  in  visual  recognition.  There 
will  be  many  examples  in  which  data,  highly  significant  for  recognition,  lie  within  the  outline.  But,  in 
general,  treatment  of  the  outline  is  a suitable  commencement  for  biometry  in  classificatory  studies. 
Measurement  loci,  as  shown  in  text-fig.  1,  show  various  degrees  of  compatibility  with  Attneave’s 
interpretation  of  visual  perception.  Some  are  located  on  outline  segments  of  low  curvature  to  which 
the  eye  gives  little  attention  (e.g.  text-fig.  1h,  i).  Others  (e.g.  text-fig.  If)  are  on  outline  segments  of 
high  curvature  that  are  probably  significant  in  object  recognition.  However,  the  use  made  of 
measurement  loci  in  most  biometrical  practice  differs  considerably  from  that  suggested  by  the 
foregoing  theory.  Biometricians  have  recorded  distances  between  loci,  whereas  theory  suggests  that  it 
is  the  position  of  loci  as  well  as  interloci  distances  that  is  important  in  perception.  A vectorial 
approach  is  indicated. 

Vectorial  data  have  been  collected  in  previous  studies  (text-fig.  2),  although  not  as  implementa- 
tions of  the  rationale  developed  above.  Examples  are  Anstey  and  Delmet  (1973)  and  Cheetham  and 
Lorenz  (1976)  on  bryozoans,  Christopher  and  Waters  (1974)  on  miospores,  Gevirtz  (1976)  and 
Pastiels  (1953)  on  bivalves,  Kaesler  and  Waters  (1972)  and  Margerie  (1977)  on  ostracods,  Scott 


text-fig.  2.  Examples  of  outline  recording,  a,  ostracod,  Eucypris,  after  Margerie 
(1971,  fig.  d).  B,  cheilostome  bryozoan,  after  Cheetham  and  Lorenz  (1976,  fig.  4). 
c,  bivalve,  Carbonicola,  after  Pastiels  (1953,  fig.  4). 


SCOTT:  OUTLINE  PROCESSING 


761 


(1976)  on  foraminifera,  and  Waters  (1977)  on  blastoids.  A common  aim  has  been  to  describe 
accurately  the  form  of  the  specimen  outline.  Although  representative  outlines  were  presented  in 
several  studies,  data  have  not  usually  been  presented  in  ways  that  assist  in  the  resolution  of  taxonomic 
problems.  For  example,  assemblies  of  outlines  (pictograms)  have,  in  the  light  of  the  previous 
discussion,  good  theoretical  support  as  effective  presentations  of  intra-sample  variation.  The 
problem  of  specimen  organization  within  the  pictogram  can  be  readily  resolved  if  outline  coordinates 
are  available. 


TECHNIQUE  AND  AN  EXAMPLE 

This  section  gives  some  simple  representations  of  outline  data  that  are  useful  in  classificatory  studies. 

Data  capture.  Text-fig.  3 summarizes  the  data  logging  and  processing  system.  The  digitizer  attached 
to  the  stereomicroscope  (Scott  1975)  was  built  to  specification  and  is  suitable  for  fossils  with  greatest 
diameters  between  0 05  mm  and  40  mm.  It  is  manually  guided  (by  movement  of  the  travelling  head) 
and  the  x,  y coordinates  of  loci  selected  by  the  operator  are  recorded  in  units  of  5-3  jum  on  paper  tape. 

Specimens  are  digitized  in  a standard  orientation.  Errors  in  orientation  are  minimized  when 
specimens  have  two  or  more  structures  that  are  small  in  relation  to  the  accuracy  of  the  measurement 
system  and  occur  in  invariant  positions.  Such  structures  are  seldom  available.  In  the  example,  the 
axial  profiles  of  the  shell  were  recorded  with  the  coiling  axis  aligned  east-west  with  reference  to  a 
cross-line  in  the  ocular  lens.  The  coiling  axis  in  foraminifera  and  similar  shells  is  not  a physical 
structure,  but  its  position  can  be  estimated  from  the  location  of  the  proloculus  (initial  chamber)  and 
umbilicus. 


TRAVELLING  A/D  CONVERTER  RAW  GRAPHICS  EDITED  PROCESSING 

MICROSCOPE  P/TAPE  ENCODER  DATA  FILE  DISPLAY  DATAFILE  SOFTWARE 


text-fig.  3.  Flow  diagram  of  data  capture,  editing,  and  processing  system.  The 
equipment  includes  a custom-built  digitizer,  Tektronix  4006  graphics  display, 
Hewlett  Packard  7202a  graphics  plotter,  and  Hewlett  Packard  2100,  Burroughs 
B6700,  and  IBM  370/168  processors. 


Editing.  Errors  due  to  mis-positioning  (backlash,  parallax,  involuntary  movement)  increase  in 
importance  as  the  size  of  the  specimen  or  structure  decreases.  Graphical  editing  of  the  recorded  x,  y 
data  is  highly  desirable.  With  batch  processing  much  can  be  done  using  lineprinter  plots  and  editing 
runs,  but  interactive  editing  with  a graphics  terminal  is  preferable.  My  equipment  displays  x,  y 
coordinates  in  order  of  recording  and  joined  by  straight  lines  (the  specimen  is  represented  as  a 
polygon).  Coordinates  may  be  inserted  or  deleted  and  the  figure  redisplayed. 

Reconstruction.  I record  about  fifty  loci  approximately  equidistant  about  the  periphery  of  the 
specimen.  There  is  no  quantitative  control  over  their  position  relative  to  the  starting-point.  Thus  the 
ith  point  on  one  specimen  is  not  necessarily  positionally  equivalent  to  the  ith  point  on  another. 
Another  consideration  is  that  only  the  obviously  spurious  coordinates  can  be  removed  by  editing. 
A residual  of  small-scale  errors  in  positioning  remains  in  the  data.  Smoothing  of  the  data  and 
interpolation  of  points  at  fixed  positions  about  the  periphery  are  performed  by  fitting  a Fourier  Series 
curve  to  each  specimen.  An  angular  expansion  of  the  radius  about  the  specimen  centroid  is  applied 
(Ehrlich  and  Weinberg  1970).  Radii  are  interpolated  at  10°  intervals  using  15  harmonics.  This 
produces  mild  smoothing.  Note  that  this  expansion  is  suitable  only  for  generally  convex  figures  in 
which  radii  are  single- valued.  All  subsequent  processing  uses  the  file  of  interpolated  radii. 


762 


PALAEONTOLOGY,  VOLUME  23 


Variation  in  Globorotalia  puncticulata  sphericomiozea.  Referred  to  this  upper  Miocene-lower 
Pliocene  planktonic  foraminiferal  taxon  are  New  Zealand  populations  that  are  intermediate  in 
morphology  and  stratigraphic  position  between  Globorotalia  miozea  conoidea  Walters  and  G. 
puncticulata  puncticulata  (Deshayes).  In  axial  orientation  G.  miozea  conoidea  is  weakly  conical  with 
the  base  formed  by  the  flattish  spiral  of  the  early  whorls  and  the  cone  by  the  ventrally  extended 
chambers  of  the  last  whorl  (for  terminology  see  text-fig.  4).  The  keel  at  the  shell  margin  is  well  defined 
on  the  last-formed  chamber  but  is  usually  buried  by  secondary  calcification  on  earlier  chambers.  The 
form  of  the  shell  in  G.  puncticulata  puncticulata  is  globose,  rather  than  conical.  This  is  produced  by 
moderate  inflation  of  chambers.  Straight-line  segments  of  the  chamber  outline  are  replaced  by  gentle 
curves.  There  is  no  keel.  At  some  horizons,  some  specimens  of  G.  puncticulata  sphericomiozea  have 
the  axial  form  of  the  ancestral  G.  miozea  conoidea  (and  its  variant  G.  conomiozea  Kennett).  Others 
anticipate  the  shape  of  G.  puncticulata  puncticulata.  Blow  (1969  p.  361)  suggested  that  such  samples 
represented  a mixture  of  two  taxa  on  the  hypothesis  that  keels,  once  evolved,  are  thereafter  retained 
in  phylogeny.  He  rejected  the  idea  of  populations  in  which  some  specimens  possessed  a keel  and 
others  did  not.  Although  there  is  no  theoretical  support  for  the  permanency  of  a structure,  Blow’s 
suggestion  about  mixed  samples  warrants  study  because  Kennett  (1977)  showed  that  there  was 
marked  deterioration  in  climate  in  the  New  Zealand  region  in  the  uppermost  Miocene,  about  the 
stratigraphic  position  of  G.  puncticulata  sphericomiozea.  Changes  in  the  distribution  of  planktonic 
taxa  in  response  to  shifts  in  watermasses  and  the  appearance  of  migrants  are  to  be  expected  in  such  a 
regime.  To  assess  Blow’s  idea,  the  systematist  needs  to  examine  intra-sample  variation.  Is  it 
continuous?  Can  sub-sample  clusters  be  detected?  Here,  the  axial  outline  of  the  shell  is  examined. 
This  profile  provides  information  on  the  shape  of  chambers  near  the  location  of  the  keel  at  the  shell 
periphery.  The  topics  considered  are  the  construction  of  a typical  outline,  and  the  pictorial 
representation  of  within-sample  variation. 


text-fig.  4.  Histograms  show  distributions  of  radii  at  20°  intervals  about 
centroids  of  fifty  specimens  of  Globorotalia  puncticulata  sphericomiozea  Walters 
from  P29/f55,  Blind  River,  New  Zealand.  The  polygonal  outline  is  formed  from 
the  mean  lengths  of  radii  spaced  at  10°  intervals. 


SCOTT:  OUTLINE  PROCESSING 


763 


Outline  representations.  The  distributions  of  radii  (text-fig.  4)  about  the  centroid  of  the  axial  outline  of 
fifty  specimens  from  P29/f55  Blind  River  (close  to  sample  32  in  Kennett  and  Watkins  1974),  show 
some  variation  in  kurtosis  but  tend  to  be  unimodal.  The  outline  in  the  centre  of  text-fig.  4 is  drawn 
from  mean  values  of  the  thirty-six  radii  and  reflects  common  features  in  the  sample  outlines  shown  in 
text-fig.  5.  Gentle  doming  in  the  vicinity  of  the  spire,  rapid  change  in  curvature  of  the  outline  of  the 
nth  chamber  at  the  site  of  the  keel,  and  ventral  extension  of  chambers  are  features  of  most  of  the 
outlines  in  text-fig.  5 that  are  also  apparent  in  the  sample  mean  outline. 


-30  | PCA  Q 1 


text-fig.  5.  Plot  of  sample  from  P29/f55  (fifty  individuals)  on  two  largest 
principal  component  axes  (dispersion  matrix,  thirty-six  radii  as  deviations  from 
means).  PCA  1 and  PCA  2 represent  81%  and  6%  of  sample  variance.  Location 
of  axial  outlines  of  specimens  is  related  to  their  position  in  the  plot  (objectively 
defined  pictogram).  Dotted  lines  show  three-cluster  division  of  sample  using  the 
non-hierarchical  clustering  algorithm  (sum  of  squares  criterion)  in  GENSTAT 
(statistical  package  produced  by  Rothamsted  Experimental  Station)  and  dashed 
line  is  the  two-cluster  partition.  This  algorithm  transfers  specimens  between 
clusters  to  improve  the  criterion  but  a global  optimum  is  not  necessarily  reached. 


However,  use  of  the  sample  mean  outline  as  a representative  form  in  comparisons  among  taxa  is 
contingent  on  negligible  shape  change  within  the  sample  size  range.  If  allometry  is  marked,  the 
sample  mean  outline  may  be  quite  unrepresentative,  not  corresponding  with  the  shape  of  any 
specimen.  Size-related  changes  in  shape  complicate  taxonomic  recognition  and  may  require  special 
study.  Brower  and  Veinus  (1978)  discussed  an  approach  suitable  for  vectorial  data.  In  the  example, 
mean  outlines  for  five  size-defined  subsamples  (text-fig.  6)  are  similar,  and  even  specimens  from  the 
extreme  size  classes  show  close  resemblance,  although  there  is  a modest  radial  extension  of  the  outline 


764 


PALAEONTOLOGY,  VOLUME  23 


in  the  vicinity  of  the  (n-2)th  chamber  of  the  largest  specimens  (text-fig.  6 centre).  I conclude  that  size- 
related  shape  changes  within  the  material  do  not  greatly  affect  the  use  of  the  sample  mean  outline  as  a 
representative  form. 

There  is  a minority  of  specimens  (e.g.  16,  23,  33  in  text-fig.  5)  in  which  spiral  and  ventral  segments 
of  the  outline  of  the  nth  chamber  form  a rounded  rather  than  an  angular  junction  (70-90°  radii  in 
text-fig.  4).  In  this  respect  they  resemble  G.  puncticulata  puncticulata.  Do  they  form  an  identifiable 
subsample?  A quantitative  or  metric  version  of  the  pictogram  (text-fig.  5),  in  which  outlines  are 
referred  to  specimen  positions  on  a principal  component  plot,  shows  that  such  specimens  are 
scattered  through  the  sample.  Thus  specimens  12  and  45  lie  at  opposite  ends  of  the  distribution  along 
PCA  1 which  represents  much  of  the  intra-sample  variation  in  outline  size.  PCA  2 reflects  variation  in 
the  degree  of  ventral  inflation  of  the  outline.  Again,  there  are  specimens  (e.g.  12,  16)  that  show 
considerable  difference  in  ventral  inflation  yet  have  rounded  peripheries. 


text-fig.  6.  Histogram  shows  distribution  of  area  enclosed  by  outlines  (axial 
profile)  of  fifty  specimens  from  P29/f55.  Area  is  taken  as  a natural  measure  of 
size.  The  superimposed  outlines  used  subsamples  based  on  the  histogram 
intervals.  Outlines  at  right  were  formed  by  ranking  the  fifty  specimens  by 
area  and  dividing  them  into  five  equal  subsamples. 


If  size  can  be  neglected  in  a taxonomic  judgement  it  is  useful  to  examine  a representation  in  which  it 
is  held  constant  (text-fig.  7).  Much  of  the  arrangement  of  text-fig.  5 is  preserved  but  there  are  several 
displacements  that  clarify  shape  similarities.  For  example,  specimen  1 (low  spire,  weak  axial 
inflation)  lies  on  the  periphery  of  the  scatter  in  text-fig.  7 whereas  in  text-fig.  5 it  lies  between 
specimens  3 and  13.  Specimens  33  and  45  are  dissimilar  in  shape  but  their  common  size  causes  their 
close  proximity  in  text-fig.  5.  They  are  widely  separated  in  text-fig.  7.  A group  of  specimens  with  weak 
axial  inflation  and  a slight  dome  representing  the  early  chambers  (e.g.  specimens  1 1, 12, 17,  32,  37, 40) 
are  in  closer  proximity  in  text-fig.  7 than  in  text- fig.  5. 

Distinct  clusters  are  not  obvious  in  text-figs.  5 and  7.  This  impression  is  supported  by  the  intra- 
sample divisions  produced  by  a non-hierarchical  clustering  algorithm  in  GENSTAT.  Large 
specimens  are  isolated  by  the  procedure  using  raw  data  (text-fig.  5,  3-cluster  partition)  but  2-cluster 
partitions  using  either  raw  or  size-standardized  data  separate  specimens  that  are  similar  in  shape  and 
in  close  proximity  in  the  principal  component  plots.  The  partitions  are  placed  in  a central  location  in 
the  scatter.  This  results  from  the  fairly  uniform  distribution  of  specimens  in  the  hyperspace. 


SCOTT:  OUTLINE  PROCESSING 


765 


text-fig.  7.  Principal  component  plot  of  the  sample  from  P29/f55  using  thirty- 
six  radii  (as  deviations  from  means)  after  areas  of  outlines  were  standardized. 
Radii  were  incremented/decremented  iteratively  until  the  area  of  each  polygonal 
outline  fell  within  5%  of  an  arbitrary  constant,  close  to  the  mean  of  the  enclosed 
area  distribution  using  raw  data.  Axes  PCA  1 and  PCA  2 represent  30%  and  24% 
of  sample  variance  (dispersion  matrix).  The  dashed  line  is  the  location  of  the 
two-cluster  partition  produced  by  the  non-hierarchical  clustering  algorithm  in 
GENSTAT  (sum  of  squares  criterion).  Axial  outlines  of  specimens  using 
standardized  data  are  arranged  according  to  their  locations  in  the  plot. 


The  data  in  text-figs.  4-7  indicate  that  a variable  population  was  sampled,  even  when  size  is 
eliminated.  But  the  representations  show  gradations  in  form  and  the  absence  of  well-defined 
disjunctions  in  specimen  distributions.  A connection  is  not  observed  between  the  form  of  the 
periphery  of  the  nth  chamber  and  the  gross  axial  shape  of  the  shell.  These  results  assist  the  taxonomist 
to  assess  the  validity  of  G.  puncticulata  sphericomiozea  in  the  light  of  Blow’s  hypothesis.  Inter-sample 
comparisons  may  also  be  useful.  Text-fig.  8,  for  example,  indicates  the  changes  in  axial  form  between 
G.  puncticulata  sphericomiozea  and  G.  puncticulata  puncticulata  much  more  explicitly  than  do  direct 
comparisons  of  specimen  suites.  In  the  latter  the  outline  of  the  nth  chamber  about  the  30-70°  segment 
(see  text-fig.  4 for  locations)  is  raised  relative  to  the  equivalent  segment  in  G.  puncticulata 
sphericomiozea.  This  occurs  throughout  the  size  range  sampled.  But  in  the  1 10-150°  segment  of  the 
nth  chamber,  inflation  relative  to  G.  puncticulata  sphericomiozea  is  marked  only  in  larger  specimens. 
The  study  of  the  transformation  in  shape  between  the  taxa  leads  to  techniques  reviewed  by  Bookstein 
(1977). 


766 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  8.  Inter-sample  comparisons  of  outlines.  Location  of  samples  P29/ 
f55  and  P29/f71  in  the  Blind  River  sequence,  scanning  electron  micrographs 
of  random  specimens  of  Globorotalia  puncticulata  puncticulata  (Deshayes) 
and  G.  puncticulata  sphericomiozea  Walters,  and  superimposed  outlines  from 
the  samples. 


CONCLUSION 

I do  not  contend  that  biometric  studies  using  ad  hoc  variates  should  be  abandoned.  Rather,  I suggest 
that  analyses  with  these  variates  usually  do  not  integrate  easily  with  qualitative  assessment  of  form. 
Generally,  they  provide  an  inadequate  representation  of  the  outline  and  may  include  measurement 
loci  not  significant  in  visual  recognition.  Vector  relationships  between  measurements  are  entirely 
omitted  yet  are  essential  in  object  identification.  By  processing  the  coordinates  of  outlines,  a 
quantitative  study  provides  information  that  is  easily  and  directly  related  to  the  material  posing  a 
classificatory  problem,  and  amenable  to  statistical  testing.  Of  course,  outline  data  may  also  contain 
significant  functional  information.  For  example,  the  form  of  the  shell  of  an  infaunal  burrower  is  likely 
to  show  adaptations  to  the  mechanism  of  movement. 

Representation  of  outlines  by  polar  coordinates  requires  large  sets  of  data  that  may  cause 
housekeeping  problems  on  small  computers.  There  is  commonly  some  redundancy  in  the  variate  set 
(dispersion  matrices  less  than  full  rank)  and  a more  parsimonious  set  is  possible.  However,  the  set 
provides  directly  a polygonal  representation  of  form  which  is  easy  to  manipulate  (magnification, 
rotation,  reflection)  and  from  which  image  descriptors  (Rink  1 976)  and  ad  hoc  variates  can  be  derived 
readily.  The  verbal  descriptors  of  Riedel  (1978)  are  less  exact  and  less  suitable  for  simple  graphical 
reconstructions  and  manipulations.  The  techniques  of  numerical  taxonomy  and  automated 
identification  (Sneath  1979)  usually  operate  with  character  states,  selected  by  the  investigator,  and  do 
not  provide  shape  representations  at  the  basic  population  level. 

Outlines  are  rich  in  information  for  the  taxonomist.  That  is  why  they  should  be  used  in  biometry. 
Nevertheless,  they  are  only  a point  of  departure.  Systems  that  process  all  pictorial  information  from  a 
specimen  suite  in  various  orientations  offer  the  prospect  of  much  more  sophisticated  assistance  to  the 
taxonomist. 


SCOTT:  OUTLINE  PROCESSING 


767 


Acknowledgements.  I am  grateful  to  A.  H.  Cheetham,  B.  W.  Hayward,  and  N.  de  B.  Hornibrook  for  reviewing 
a draft  of  this  paper. 


REFERENCES 

anstey,  R.  l.  and  delmet,  D.  A.  1973.  Fourier  analysis  of  zooecial  shapes  in  fossil  tubular  bryozoans.  Bull.  geol. 
Soc.  Am.  84,  1753-1764. 

attneave,  F.  1954.  Some  informational  aspects  of  visual  perception.  Psychol.  Rev.  61,  183-193. 
berthou,  P.  Y.,  brower,  J.  c.  and  reyment,  R.  A.  1975.  Morphometrical  study  of  Choffat’s  vascoceratids  from 
Portugal.  Bull.  geol.  Instn  Univ.  Uppsala  N.s.  6,  73-83. 

blow,  w.  H.  1969.  Late  middle  Eocene  to  Recent  planktonic  foraminiferal  biostratigraphy.  In  bronnimann,  p. 
and  renz,  H.  H.  (eds.).  Proceedings  of  the  First  International  Conference  on  Planktonic  Microfossils,  Brill, 
Leiden,  1,  199-421. 

bookstein,  F.  L.  1977.  The  study  of  shape  transformation  after  D’Arcy  Thompson.  Math.  Biosci.  34,  177-219. 
brower,  j.  c.  and  veinus,  j.  1978.  Multivariate  analysis  of  allometry  using  point  coordinates.  J.  Paleont.  52, 
1037-1053. 

cadot,  h.  m.  and  kaesler,  r.  l.  1973.  Variation  of  carapace  morphology  of  bairdiacean  and  cytheracean 
Ostracoda.  Paleont.  Contr.  Univ.  Kans.  61,  1-10. 

cheetham,  A.  h.  and  lorenz,  d.  m.  1976.  A vector  approach  to  size  and  shape  comparisons  among  zooids  in 
cheilostome  bryozoans.  Smithson.  Contr.  Paleobiol.  29,  1-55. 

Christopher,  r.  a.  and  waters,  j.  a.  1974.  Fourier  series  as  a quantitative  descriptor  of  miospore  shape. 
J.  Paleont.  48,  697-709. 

ehrlich,  r.  and  weinberg,  b.  1970.  An  exact  method  for  characterization  of  grain  shape.  J.  sedim.  Petrol.  40, 
205-212. 

fisher,  w.  l.,  rodda,  p.  u.  and  dietrich,  j.  w.  1964.  Evolution  of  the  Athleta  petrosa  stock  (Eocene,  Gastropoda) 
of  Texas.  Univ.  Tex.  Pubis  6413,  1-117. 

gevirtz,  J.  L.  1976.  Fourier  analysis  of  bivalve  outlines:  implications  on  evolution  and  autecology.  J.  Int.  Ass. 
Math.  Geol.  8,  151-163. 

hallam,  A.  and  gould,  s.  J.  1975.  The  evolution  of  British  and  American  middle  and  upper  Jurassic  Gryphaea: 
a biometric  study.  Proc.  R.  Soc.  Lond.  B 189,  511-542. 

kaesler,  r.  l.  and  waters,  j.  a.  1972.  Fourier  analysis  of  the  ostracod  margin.  Bull.  geol.  Soc.  Am.  83, 
1169-1178. 

kennett,  j.  p.  1977.  Cenozoic  evolution  of  Antarctic  glaciation,  the  circum-Antarctic  Ocean,  and  their  impact 
on  global  paleoceanography.  J.  geophys.  Res.  82,  3843-3860. 

— and  watkins,  n.  d.  1974.  Late  Miocene-early  Pliocene  paleomagnetic  stratigraphy,  paleoclimatology,  and 
biostratigraphy  in  New  Zealand.  Bull.  geol.  Soc.  Am.  85,  1385-1398. 

koffka,  K.  1935.  Principles  of  gestalt  psychology.  Routledge  & Kegan  Paul,  London,  720  pp. 
margerie,  p.  1971.  Essai  de  ‘quantification’  du  contour  des  ostracodes  a l’occasion  de  la  description  d’une 
nouvelle  espece  de  Cypridinae  du  Marinesien  du  bassin  de  Paris.  Revue  Micropaleont.  14,  227-234. 

— 1977.  Complement  a deux  articles  anterieurs  (1969-1972)  relatifs  a l’utilisation  de  la  statistique  pour  la 
description  des  ostracodes.  Ibid.  20,  44-48. 

marr,  D.  1976.  Early  processing  of  visual  information.  Phil.  Trans.  R.  Soc.  Lond.  B 275,  483-524. 
mateer,  N.  J.  1976.  A statistical  study  of  the  genus  Pterodactylus.  Bull.  geol.  Instn.  Univ.  Uppsala,  n.s.  6,  97-105. 
melville,  R.  1978.  On  the  discrimination  of  species  in  hybrid  swarms  with  special  reference  to  Ulmus  and  the 
nomenclature  of  U.  minor  Mill,  and  U.  carpinifolia  Gled.  Taxon,  27,  345-351. 
olson,  E.  c.  1953.  Integrating  factors  in  amphibian  skulls.  J.  Geol.  61,  557-568. 

— and  miller,  r.  l.  1958.  Morphological  Integration,  Univ.  Chicago  Press,  Chicago,  317  pp. 

pastiels,  A.  1953.  Etude  biometrique  des  Anthracosiidae  du  Westphalien  A de  la  Belgique.  Pubis  Ass.  Etude 
Paleont.  Stratigr.  houill.  16,  1-56. 

reyment,  R.  a.  1963.  Notes  on  the  description  of  post-Paleozoic  fossil  ostracods.  J.  Paleont.  37,  682-687. 
riedel,  w.  r.  1978.  Systems  of  morphologic  descriptors  in  paleontology.  Ibid.  52,  1-7. 

rink,  M.  1976.  A computerized  quantitative  image  analysis  procedure  for  investigating  features  and  an  adapted 
image  process.  J.  Microsc.  107,  267-286. 

rowell,  a.  J.  1966.  Revision  of  some  Cambrian  and  Ordovician  inarticulate  brachiopods.  Paleont.  Contr.  Univ. 
Kans.  7,  1-36. 

scott,  G.  h.  1972.  The  relationship  between  the  Miocene  Foraminiferida  Globorotalia  miozea  miozea  and  G. 
praemenardii.  Micropaleontology,  18,  81-93. 


768 


PALAEONTOLOGY,  VOLUME  23 


SCOTT,  G.  H.  1975.  An  automated  coordinate  recorder  for  biometry.  Lethaia,  8, 49-52. 

— 1976.  Estimation  of  ancestry  in  planktonic  foraminifera:  Globoquadrina  dehiscens.  N.Z.  J.  Geol.  Geophys. 
19,311-325. 

sellberg,  B.  and  kjellstrom,  G.  1975.  Geometric  characterization  of  acritarchs  belonging  to  the  genus 
Veryhachium.  Neues  Jb.  Geol.  Palaont.,  Mh.  1975,  310-314. 

Sergeyeva,  s.  p.  1975.  Orientation,  morphological  terminology  and  measurements  of  simple  conodonts.  Paleont. 
J.  9,  385-391. 

shaw,  a.  b.  1957.  Quantitative  trilobite  studies  II.  Measurement  of  the  dorsal  shell  of  non-agnostidean  trilobites. 
J.  Paleont.  31,  193-207. 

SIMPSON,  G.  G.,  ROE,  a.  and  LEWONTIN,  R.  c.  1960.  Quantitative  Zoology,  Harcourt  Brace,  New  York,  440  pp. 
sneath,  p.  h.  a.  1979.  Numerical  taxonomy  and  automated  identification:  some  implications  for  Geology. 
Computers  and  Geosciences,  5,  41-46. 

temple,  j.  t.  1975.  Standardization  of  trilobite  orientation  and  measurement.  Fossils  and  Strata,  4,  461-467. 
waters,  J.  A.  1977.  Quantification  of  shape  by  use  of  Fourier  analysis:  the  Mississippian  blastoid  genus 
Pentremites.  Paleobiology,  3,  288-299. 


Typescript  received  13  September  1979 
Revised  typescript  received  26  November  1979 


G.  H.  SCOTT 

N.Z.  Geological  Survey 
P.O.  Box  30368 
Lower  Hutt 
New  Zealand 


HI  A TELLA  — A JURASSIC  BIVALVE  SQUATTER? 

by  SIMON  R.  A.  KELLY 


Abstract.  English  late  Jurassic  (Middle  Volgian)  Hiatella  occur  in  two  habitats;  firstly,  as  simple  byssal  nestlers 
on  local  hard  substrates  and,  secondly,  within  Gastrochaenolites- type  borings  penetrating  hard  substrates.  Most 
Hiatella  occupy  borings  that  they  did  not  originally  construct  themselves,  although  ancestors  as  well  as  other 
bivalve  genera  could  have  been  responsible.  The  morphology  of  the  Mesozoic  Hiatella  is  compared  briefly  with 
modern  species  which  occur  around  the  British  Isles  and  which  include  both  boring  and  nestling  forms. 
A sequence  of  events  is  postulated  for  the  formation  of  the  Basal  phosphatized  Nodule  Bed  of  the  Spilsby 
Sandstone  in  Lincolnshire,  and  a palaeoenvironmental  model  is  suggested  for  the  East  Midlands  Shelf  in  Middle 
Volgian  times. 


The  borings  made  by  bivalves  into  hard  substrates  have  been  the  subject  of  considerable  attention 
from  both  zoologists  and  palaeontologists  and  there  are  many  important  articles  in  the  publications 
edited  by  Clapp  and  Kenk  (1963),  Crimes  and  Harper  (1970,  1977),  and  Frey  (1975).  Unlike  most 
trace  fossils,  borings  of  bivalves  may  commonly  contain  the  skeletal  remains  of  their  occupants. 
However,  caution  is  necessary  in  recognizing  whether  the  occupant  is  primary,  i.e.  the  organism 
which  originally  constructed  the  boring,  or  whether  it  is  secondary  and  is  effectively  a squatter  in  the 
vacated  domicile.  There  is  ample  evidence  of  modern  bivalves  reoccupying  vacant  borings,  largely 
those  of  pholads,  but  including  (updated  names)  Tresus,  Petricola , Macoma,  and  Irus  (Evans  1967); 
Kellia  and  Notirus  (Stevenson  1946);  Tapes , Cumingia , Kelli  a,  Diplodontct , Endodesma , and  My  ti  fits 
(Barrows  1917);  Modiola,  Scaphula,  and  Corbula  in  Mar  tesla  borings  in  brickwork  (Annandale 
1923);  Idasola  in  borings  of  Teredo  in  wood  (Jensen  1912).  Kiihnelt  (1933,  1951)  recorded  Ungulina, 
Montacuta,  Lepton , Coralliophaga,  Trapezium , Venerupis , Sphenia,  Perna,  Lyonsia , Petricola , and 
Hiatella , all  of  which  are  deformed  to  some  degree  to  fit  the  borings  in  which  they  occur.  Some 
bivalves  like  Hiatella  and  Petricola  (Yonge  1958;  Hunter  1949)  may  either  bore  into  hard  substrates 
or  nestle  epibyssally.  Bivalve  borings  in  turn  may  be  reinfested  by  other  phyla,  e.g.  hydroids  and 
bryozoa  described  by  Evans  (1949),  surviving  in  the  wet  microenvironments  of  the  vacant  borings  in 
the  intertidal  zone.  Warme  (1970)  noted  that  abandoned  borings  may  be  modified  and  deepened  by 
nestling  bivalves,  gastropods,  polychaetes,  arthropods,  etc. 

Records  of  fossil  bivalves  reoccupying  vacant  borings  are  much  less  common.  Masuda  (1968) 
noted  Barbatia,  Irus,  and  Phlyctiderma  in  partially  eroded  Miocene  borings.  Itoigawa  (1963) 
recorded  the  borings  of  Miocene  Parapholas  which  were  subsequently  infilled  by  sediment,  and  then 
burrowed  by  Lutraria  before  consolidation.  Kennedy  and  Klinger  (1972)  discussed  a number  of 
encrusting  and  nestling  organisms  occupying  borings  constructed  by  a Cretaceous  mytilid;  these 
include  serpulids,  a bryozoan,  ostreids,  and  Barbatia.  Jurassic  Hiatella  has  been  recognized  only 
rarely.  Eudes-Deslonchamps  (1838)  ascribed  two  species  from  the  Middle  Jurassic  of  Normandy  to 
Saxicava  (a  junior  synonym  of  Hiatella),  and  Chavan  (1952)  introduced  the  genus  Pseudosaxicava 
for  a Lower  Kimmeridgian  species  from  the  same  area,  and  this  name  is  placed  as  a subgenus  of 
Hiatella  by  Keen  (in  Moore  1969).  From  England,  Cox  (1929)  described  ‘ Area ' foetida  from  the 
Portland  Sand  and  Hartwell  Clay  (Middle  Volgian).  This  latter  species  is  conspecific  with  other 
material  described  here  from  the  Middle  Volgian.  The  updated  name  of  this  species  is  Hiatella 
(Pseudosaxicava)  foetida  (Cox  1929). 

There  has  been  little  ecological  information  associated  with  these  early  records,  though  Eudes- 
Deslongchamps  noted  that  his  Middle  Jurassic  examples  were  associated  with  borings  into  corals  and 
bivalve  shells.  The  description  here  of  specimens  from  the  English  Middle  Volgian  adds  significantly 

[Palaeontology,  Vol.  23,  Part  4,  1980,  pp.  769-781,  pi.  96.| 


770 


PALAEONTOLOGY,  VOLUME  23 


to  the  paleoecology  of  Hiatella.  There  is  evidence  that  the  shell  shape  is  strongly  controlled  by  the 
substrate  to  which  it  is  attached.  There  is  little  positive  evidence  for  English  Upper  Jurassic  Hiatella 
having  been  capable  of  boring,  while  there  is  plenty  of  evidence  which  indicates  that  vacant  bivalve 
borings  were  commonly  infested  by  Hiatella  spat.  Modern  British  Hiatella  have  been  studied  by 
Hunter  (1949),  who  described  considerable  variation  in  shell  shape  which  is  closely  paralleled  by  the 
late  Jurassic  forms,  depending  largely  on  whether  they  are  boring  or  nestling.  Strauch  (1968) 
suggested  that  the  shell  length  of  Recent  Hiatella  was  inversely  related  to  the  winter  minimum  water 
temperature  and  consequently  was  useful  in  estimation  of  Cenozoic  palaeotemperatures.  However, 
this  is  partially  doubted  by  Rowland  and  Hopkins  (1971)  who  believe  that  there  is  a more  complex 
situation  and  that  size  is  controlled  more  by  mode  of  life  in  each  population. 

STRATIGRAPHY 

The  specimens  used  in  this  study  are  all  from  the  Middle  Volgian  (equivalent  to  the  upper  part  of  the 
Upper  Kimmeridgian  and  the  lower  part  of  the  Portlandian  of  England).  Extensive  collecting  was 
carried  out  in  the  Basal  Spilsby  Nodule  Bed  in  a sand  pit,  now  bulldozed,  on  Nettleton  Hill, 
Lincolnshire  (TF  108989)  (see  text-fig.  1 for  localities).  Although  in  situ  collecting  is  no  longer 
possible  at  this  site,  the  hillside  about  200  m to  the  north  provides  much  weathered-out  loose  material 
from  the  same  horizon.  The  collections  made  from  this  horizon  have  been  deposited  with  the  Institute 
of  Geological  Sciences,  London  (IGS).  Casey  (1973)  referred  this  bed  to  the  Titanites  giganteus  Zone. 
The  status  of  this  zone  in  Lincolnshire  is  not  clear  since  Wimbledon  and  Cope  (1978)  have  completely 
revised  the  zonal  sequence  in  southern  England.  However,  it  is  possible  that  the  fauna  of  this  bed  may 
represent  several  zones  as  repeated  phases  of  phosphatization  can  be  recognized  and  the  ammonites 
(all  phosphatized)  belong  to  the  genera  Crendonites,  Epilaugeites,  Kerberites,  and  Pavlovia  (R.  Casey 
pers.  comm.).  The  Basal  Spilsby  Nodule  Bed  rests  upon  eroded,  plastic  blue-grey  Kimmeridge  Clay 
with  occasional  cementstones  up  to  0-2  m thick  and  containing  Pectinatites  of  Lower  Volgian  age. 
The  nodule  bed  itself  is  about  0-2  m thick  and  is  composed  of  brown  and  blackened  phosphatized 
concretions  up  to  0-2  m in  diameter,  but  commonly  10-30  mm,  together  with  small  lyditic  pebbles  set 
in  a dark,  glauconitic  silty  sand.  Many  of  the  concretions  show  compound  structure  and  are 
commonly  abraded,  showing  signs  of  bioerosion,  e.g.  flask-shaped  borings  attributable  to  bivalves 


text-fig.  1 . Sketch  map  of  the  distribution  of 
Middle  Volgian  strata  in  England,  with  loca- 
tions of  sites  where  Hiatella  has  been  obtained. 


KELLY:  JURASSIC  BIVALVE  SQUATTER? 


77: 


and  grazing  trails  probably  caused  by  gastropods.  A rich  fauna,  especially  of  bivalves,  has  been 
obtained  from  this  bed  (Kelly  1977).  The  preservation  of  the  fauna  is  normally  as  hollow 
phosphatized  moulds,  with  internal  moulds  of  bivalves  and  of  parts  of  ammonites  making  up  a high 
proportion  of  the  nodules  of  the  bed.  Above  the  nodule  bed  lies  0-6  m of  poorly  consolidated 
glauconitic  silty  sand,  the  base  of  which  is  pale  coloured,  becoming  brown  (ferruginous)  near  the 
centre  and  grey  at  the  top,  and  which  contains  unidentified,  partly  phosphatized,  ammonites. 

Similar  phosphatized  material  with  Hiatella  occurs  in  the  base  of  the  Lower  Greensand  at  Upware, 
Potton,  and  Brickhill,  and  is  preserved  in  the  Sedgwick  Museum,  Cambridge.  Although  these 
specimens  are  mixed  with  other  phosphatized  material  ranging  from  Oxfordian  to  Aptian  in  age,  they 
occur  with  ammonites,  a large  proportion  of  which  are  of  Middle  Volgian  age  and  they  are 
undoubtedly  of  the  same  age.  Unphosphatized  Hiatella  occur  in  the  Hartwell  Clay  of  Buckingham- 
shire and  the  Swindon  Clay  of  Wiltshire  (both  of  Pavlovia  pallasioides  Zone)  and  are  preserved  in  the 
British  Museum  (Natural  History),  the  Institute  of  Geological  Sciences,  and  the  Sedgwick  Museum, 
Cambridge  (e.g.  PI.  96,  figs.  15, 16).  From  an  unspecified  horizon  in  the  Portland  Sand  of  Hounstout, 
Dorset  (Waddington  Collection,  untraced),  two  specimens  were  figured  as  ‘ Area'  foetida  sp.  nov.  by 
Cox  (1929,  pi.  1,  figs.  2,  3).  These  specimens  are  likely  to  have  come  from  the  horizons  recorded 
by  Arkell  (1935,  p.  310),  who  listed  Parallelodon  (Beshausenia)  foetidum  from  the  White  Cementstone 
and  Bed  1 1 of  the  Emmit  Hill  Marls.  In  the  latter  horizon  Arkell  noted  that  another  more  elongate 
species  of  the  genus  was  also  present.  Hiatella  has  also  been  collected  recently  from  borings  in  the 
upper  part  of  the  Portland  Limestone  on  the  Isle  of  Portland.  It  is  interesting  to  note  that  Woodward 
(1851-1856)  recorded  modern  Hiatella  actively  attacking  the  Portland  Stone  breakwater  at 
Plymouth,  which  perhaps  even  makes  possible  the  reoccupation  of  Jurassic  borings  after  some  135 
million  years. 


DESCRIPTION  OF  BORINGS  AND  THE  HIATELLA 

In  the  Basal  Spilsby  Nodule  Bed,  Hiatella  was  collected  both  from  within  flask-shaped  borings  and 
independently  of  the  borings.  These  two  types  appear  to  be  morphologically  distinct  and  are 
therefore  described  separately,  although  it  is  possible  to  find  intermediate  forms.  As  much  of  the 
discussion  in  this  paper  centres  around  the  occurrence  of  Hiatella  in  the  borings,  these  structures  are 
described  first,  followed  by  details  of  the  shell  shape  in  both  the  boring  and  the  non-boring  habitat. 

The  borings.  The  Basal  Spilsby  nodules  contain  several  types  of  borings  of  which  the  most 
conspicuous  are  flask-shaped  cavities  or  their  phosphatized  infillings,  commonly  up  to  30  mm  in 
length  (text-fig.  2 a-c).  The  flask  is  circular  in  cross-section  (text-fig.  3c)  with  a maximum  diameter  of 
13  mm.  The  constricted  neck  reaches  5 mm  diameter  and  is  circular  except  near  the  aperture,  where  it 
becomes  slightly  oval  and  weakly  flared  (PI.  96,  fig.  23).  Oblique  sections  through  the  flask  may  be 


text-fig.  2.  Camera  lucida  drawings  of  phosphatized  infillings  of  Gastrochaenolites  borings  in  Basal  Spilsby 
Nodule  Bed,  Nettleton,  Lincolnshire,  S.  R.  A.  Kelly  Collection  IGS:  A,  Zu2229;  B,  Zu2230;  C,  Zu2228;  D, 

Zu2231;  E,  Zu2232. 


772 


PALAEONTOLOGY,  VOLUME  23 


text-fig.  3.  Camera  lucida  drawings  of  polished  sections  through  phosphatized  compound 
nodules  of  the  Basal  Spilsby  Nodule  Bed  showing  Gastrochaenolites  borings;  some  (figs.  A 
and  b)  show  Hiatella  in  sites  within  the  borings.  Nettleton,  Lincolnshire.  S.  R.  A.  Kelly 
Collection,  IGS:  A,  Zu2237;  B,  Zu2223;  C,  Zu2238;  D,  Zu2236. 

pear-shaped  (text-fig.  3a).  A complete  longitudinal  section  through  an  infilled  boring  is  shown  in  text- 
fig.  3 d.  The  borings  are  preserved  as  hollows  penetrating  the  already  phosphatized  nodules.  They  may 
penetrate  both  nodules  and  phosphatized  matrix  alike  without  break,  which  indicates  that  the 
substrate  was  evenly  lithified  despite  an  apparent  heterogeneous  nature.  Each  phase  of  phosphatiza- 
tion  can  be  distinguished  by  a darkened  outer  margin.  Absence  of  crushed  or  distorted  borings  also 
shows  that  the  substrate  was  completely  lithified.  Borings  may  not  be  perfectly  straight  but  may  have 
bent  necks  (text-fig.  2b).  These  are  presumably  due  to  the  original  boring  organism  modifying  the 
direction  of  boring  because  of  unsuitable  substrate  or  of  crowding  by  other  individuals. 
Interpenetrating  borings  also  occur  (text-figs.  2c,  e).  The  first-formed  boring  appears  to  be  infilled 
and  phosphatized  before  being  cut  across  by  a second  boring. 

Although  the  substrate  of  these  borings  is  normally  a phosphatized  nodule,  one  particular  example 
shows  a large  piece  of  reptilian  bone  which  has  been  attacked.  The  upper  surface  of  the  bone  (PI.  96, 
fig.  23)  shows  that  little  erosion  has  taken  place  since  the  original  construction  of  the  borings  as  the 
openings  are  still  oval.  The  whole  flasks  can  be  seen  in  Plate  96,  fig.  24,  together  with  a specimen  of 


KELLY:  JURASSIC  BIVALVE  SQUATTER? 


773 


Hiatella  in  situ  in  one  of  them.  The  bone  must  have  been  buried  before  abrasion  destroyed  the  oval 
necks  of  the  borings.  Another  specimen,  not  figured,  shows  a boring  penetrating  an  icthyosaurian 
vertebra.  In  contrast,  Plate  96,  fig.  18  shows  part  of  a phosphatized  nodule  that  has  been  bored  and 
subsequently  abraded  so  deeply  prior  to  final  burial  that  only  rounded  bases  of  the  deepest  part  of  the 
borings  remain  visible.  The  borings  are  normally  found  penetrating  nodules;  however,  during  phases 
of  reworking  the  nodules  may  become  broken  and  the  lithified  boring  infillings  become  loose.  Such 
infillings  may  be  found  reworked  into  the  sediment  as  clasts  in  the  manner  described  by  Radwanski 
(1977). 

These  borings  correspond  closely  to  the  ichnogenus  Gastrochaenolites  Leymerie  (1842),  originally 
described  from  the  Calcaire  a Spatangues,  Neocomian,  Aube,  France.  This  name  was  not  included  in 
the  Treatise  (Hantzschel  1975).  Leymerie  clearly  described  Gastrochaenolites  as  a boring  in  rock 
which  was  found  in  association  with  Gastrochaena  dilatata  Deshayes.  It  is  distinguished  from 
Teredolites  Leymerie  (1842)  which  penetrated  wood  and  is  more  evenly  tapered  along  its  length. 
Bromley  (1972)  placed  both  Gastrochaenolites  and  Teredolites  with  the  more  recent  ichnotaxon 
Trypanites  Magdefrau  (1932),  which  Hantzschel  (1975)  restricted  to  straight-sided  tunnels  of  1 -2  mm 
width.  The  ichnogenus  Gastrochaenolites  is  retained  here  for  the  Basal  Spilsby  Nodule  Bed  borings 
until  the  taxonomy  of  these  ichnogenera  is  clarified. 

Evans  (1970)  showed  that  with  increasing  rock  hardness  the  ratio  of  the  valve  length  to  valve 
depth  decreased  for  Penitella,  and  the  weight  of  a valve  of  given  size  increased.  As  a consequence,  the 
shape  of  the  boring  also  changed,  becoming  shorter  and  broader  with  increased  hardness.  It  has  not 
yet  been  possible  to  compare  in  detail  the  borings  containing  Hiatella  from  the  Portland  Stone  in 
southern  England,  and  therefore  varied  substrates  cannot  be  compared  to  show  whether  the  hardness 
of  the  substrate  affected  the  shape  of  the  boring.  There  is  also  the  problem  of  establishing  without 
doubt  the  original  constructor  of  the  boring  and  if  several  different  bivalves  are  constructing  the 
borings  they  may  each  have  distinctive  sized  and  shaped  borings. 

Hiatella  in  the  borings.  Specimens  of  Hiatella  found  inside  Gastrochaenolites  borings  in  the  Basal 
Spilsby  Nodule  Bed  range  up  to  12  mm  in  length.  They  are  preserved  as  internal  and  external  moulds 
in  phosphorite.  The  specimens  illustrated  on  Plate  96  are  largely  casts  made  from  silicone  rubber.  The 
distinctive  features  of  these  specimens  are:  the  tendency  of  the  two  carinae  bounding  the  dorsal  and 


text-fig.  4.  Sketches  of  Recent  and  Jurassic  Hiatella  valves  to  illustrate  shell  form  in  boring  and  non-boring 
habit,  a,  Hiatella  ( Hiatella ) arctica  (Linne),  non-boring  habitat.  Recent  (after  Hunter  1949);  h,  H.  (H.)  gallicana 
(Lamarck),  boring  habitat,  Recent  (after  Hunter  1949);  c,  H.  ( Pseudosaxicava ) foetida  (Cox),  non-boring 
habitat,  Middle  Volgian;  d,  H.  ( P .)  foetida  (Cox),  boring  habitat.  Middle  Volgian. 


774 


PALAEONTOLOGY,  VOLUME  23 


ventral  margins  of  the  posterior  area  to  be  distinct  only  close  to  the  umbo  and  to  disappear  gradually 
towards  the  posterior  margin  (PL  96,  figs.  1-6,  19;  text-fig.  4 d);  the  posterior  area  tends  to  be  weakly 
inflated  and  the  comarginal  ornament  is  normally  suppressed;  the  umbones  are  usually  low;  the 
growth-lines  may  become  crowded  towards  the  commissure  and  there  is  little  trace  of  median  sulcus 
on  the  ventral  margin.  All  these  features  suggest  that  the  shell  may  be  becoming  confined  by  the  shape 
of  the  boring  in  which  it  lived.  Some  specimens,  however,  are  clearly  too  small  to  have  constructed  the 
boring  (text-fig.  3 a,  b;  Plate  96,  fig.  20)  and  also  two  individuals  have  been  found  in  the  same  boring 
(PI.  96,  fig.  22);  such  specimens  lack  features  which  indicate  confining  by  the  boring  and  tend  to  have 
fully  developed  ornament.  These  features  in  general  indicate  that  the  Hiatella  is  infesting  borings 
which  are  not  of  its  own  making. 

Gastrochaena  in  borings.  About  150  specimens  of  Hiatella  have  been  found  in  Gastrochaenolites 
borings  in  the  Basal  Spilsby  Nodule  Bed.  However,  one  rock  specimen  has  two  Gastrochaenolites 
borings  containing  the  bivalve  Gastrochaena  itself  (PI.  96,  fig.  17)  and  a single  external  mould  of  a 
right  valve  of  Gastrochaena  shown  as  a cast  in  Plate  96,  fig.  21 . Recent  Gastrochaena  sensu  stricto  is  well 
known  as  a borer  into  calcareous  substrates  in  temperate  and  tropical  regions.  It  is  distinguished 
from  Hiatella  by  its  large  anterior  pedal  gape  and  its  lack  of  external  ornament  like  carinae  and 
lamellae.  The  borings  associated  with  the  Spilsby  Gastrochaena  fit  tightly  around  the  shells  and  show 
weak  traces  of  the  calcareous  extension  tubes,  which  are  not  actually  seen  on  any  borings  associated 
with  Hiatella.  It  is  not  clear  whether  Gastrochaena  was  a precursor  to  the  Hiatella  in  the  borings  of  the 
Basal  Spilsby  Nodule  Bed,  or  whether  the  two  were  contemporaneous. 

Hiatella  independent  of  borings.  The  best-preserved  examples  of  Hiatella  found  independently  of  the 
borings  are  the  aragonitic  examples  from  the  Hartwell  and  Swindon  Clays  (PI.  96,  figs.  15,  16).  Such 
specimens  are  normally  found  as  disarticulated  valves,  while  those  from  the  Basal  Spilsby  Nodule 
Bed  are  normally  complete  internal  phosphatized  moulds  with  valves  in  occlusion  (steinkerns)  (PI. 
96,  figs.  7,  8,  11-14).  The  independent  shells  commonly  range  up  to  a larger  size  (30  mm)  than  those 
from  the  borings.  Although  the  upper  length  limit  of  30  mm  is  identical  to  the  maximum  length  of  the 
borings,  the  maximum  expected  size  of  a Hiatella  in  a boring  would  be  about  20  mm,  because  of  the 
constricted  neck  area.  Presumably  the  destruction  of  further  large  Gastrochaenolites  specimens 
would  provide  larger  Hiatella  than  the  12  mm  recorded  above.  The  shell  is  more  oval  in  cross-section; 


EXPLANATION  OF  PLATE  96 

Figs.  1-14,  19,  20,  22.  Hiatella  ( Pseudosaxicava ) foetida  (Cox).  1,  2,  cast  of  complete  individual,  IGS  Zu2216, 
2217,  x 1.  3,  4,  cast  of  complete  individual,  IGS  Zu2219,  x 1.  5,  6,  cast  of  incomplete  individual,  IGS 
Zu2222,  x 1 . 7,8,  phosphatized  steinkern,  IGS  Zu224 1 , x 1 . 9,10,  cast  of  complete  individual,  IGS  Zu22 1 8, 
2219,  2220,  x 1.  11,12,  phosphatized  steinkern  with  cast  of  some  adhering  shell,  IGS  Zu2242,  x 1.  13,  14, 
phosphatized  steinkern,  IGS  Zu2243,  x 1 . 19,  phosphatized  internal  mould  completely  fitting  within  boring, 
IGS  Zu2225,  x 1 . 20,  cast  within  boring  that  is  too  small  to  have  been  made  by  this  occupant,  IGS  Zu2234, 
2235,  x 1-5.  22,  two  phosphatized  internal  moulds  of  right  valves  representing  two  individuals  within  the 
same  boring;  Basal  Spilsby  Nodule  Bed,  Middle  Volgian,  Nettleton,  Lincolnshire. 

Figs.  15,  16.  H.  ( P .)  foetida  (Cox).  Right  valve  exterior,  IGS  Y709,  Hudleston  Collection,  x 1;  Upper 
Kimmeridge  Clay,  Pavlovia  pallasioides  Zone,  Middle  Volgian,  Swindon,  Wiltshire. 

Figs.  17,  21.  Gastrochaena  sp.  17,  individuals  with  Gastrochaenolites- type  borings,  x 1-5.  21,  cast  (seen  as 
mould  on  fig.  17)  of  left  valve,  x2.  IGS  Zu2224.  Basal  Spilsby  Nodule  Bed,  Middle  Volgian,  Nettleton, 
Lincolnshire. 

Figs.  18,  23,  24.  Gastrochaenolites  ichnosp.  18,  eroded  flask  bases,  IGS  Zu2226,  x 1-5.  23,  reptilian  bone 
showing  oval  apertures  to  flask-shaped  borings.  24,  same  specimen  in  broken  section  showing  opened  flasks 
and  an  individual  Hiatella  steinkern  in  situ  in  one,  IGS  Zu2227,  x 2.  Basal  Spilsby  Nodule  Bed,  Middle 
Volgian,  Nettleton,  Lincolnshire. 


PLATE  96 


KELLY,  Jurassic  boring  bivalves 


776 


PALAEONTOLOGY,  VOLUME  23 


the  posterior  carinae  are  distinct  throughout  their  length;  the  posterior  area  is  gently  concave; 
comarginal  lamellae  are  well  developed  on  the  posterior  area,  and  the  ventral  margin  is  usually  gently 
sulcate,  the  latter  feature  giving  the  byssate  shell  greater  stability  in  currents  (PI.  96,  figs.  9,  10;  text- 
fig.  4c).  Unfortunately  Oates  (1974),  in  his  palaeoecological  study  of  the  Hartwell  Clay,  did  not 
recognize  Hiatella,  although  the  collections  he  examined  do  contain  them,  but  they  tend  to  be 
confused  with  species  of  Grammatodon.  I believe  that  in  the  Hartwell  and  Swindon  clays  both  the 
Hiatella  and  Grammatodon  are  byssate  nestlers  and  not  shallow  infauna  as  Oates  suggested.  Both 
these  taxa  may  show  a weak  byssal  gape. 

The  non-boring  Hiatella  are  believed  to  have  been  byssally  attached  to  the  exterior  of  local  hard 
substrates  such  as  shells  of  ammonites  and  phosphatized  nodules.  Uninhibited  growth  allowed  the 
shells  to  grow  to  a greater  size  than  in  the  borings.  The  large  number  of  complete  internal  moulds  in 
the  Basal  Spilsby  Nodule  Bed,  as  opposed  to  isolated  valves,  probably  reflects  rapid  burial,  with  the 
shells  still  attached  to  the  substrate.  Early  diagenetic  phosphatization  took  place  within  the  reduced 
zone  defined  by  the  valves.  Subsequent  winnowing  and  destruction  of  the  shell  concentrated  the 
internal  moulds  together  with  other  phosphatized  debris.  A reconstruction  of  a Basal  Spilsby  Nodule 
infested  with  boring  and  non-boring  Hiatella  is  shown  in  text-fig.  5. 


text-fig.  5.  Reconstruction  of  a Basal  Spilsby  Sandstone  phosphatized  nodule,  partially  cut  away  to  illustrate 
Hiatella  ( Pseudosaxicava ) foetida  (Cox)  in  its  two  ecological  niches.  The  smaller,  more  constricted  shelled 
specimens  occupy  the  borings,  while  the  larger,  more  fully  developed  examples  are  epibyssally  attached  to  the 
exterior  of  the  nodule.  For  simplification  the  abundant  and  varied  associated  fauna  of  bivalves,  gastropods, 
brachiopods,  serpulids,  etc.  are  omitted. 


KELLY:  JURASSIC  BIVALVE  SQUATTER? 


Ill 


DISCUSSION 

Recent  Hiatella  are  byssally  attached  to  the  substrate  of  their  choice,  whether  living  epifaunally  on 
hard  substrates  or  infaunally  in  borings.  The  range  in  shape  of  the  Jurassic  shells  is  very  similar  to  that 
of  the  recent  British  species  (discussed  by  Hunter  1949),  and  there  is  little  reason  to  suspect  that  they 
lived  in  different  ways. 

Hunter  (1949)  recognized  two  recent  species;  the  first,  H.  gallicana  (Lamarck)  (text-fig.  4b)  is 
normally  found  inhabiting  borings  in  calcareous  substrates.  The  shell  shows  features  akin  to  Hiatella 
from  borings  in  the  Basal  Spilsby  Nodule  Bed,  in  particular  the  suppression  of  the  umbo,  posterior 
carinae,  and  lamellae.  The  species  is  accepted  as  a rock  borer  and  is  believed  to  bore  with  the  foot 
using  sand  grains  and  mucus  as  an  abrasive.  There  is  as  yet  no  positive  evidence  for  any  chemical 
secretion  being  used  as  in  the  calcium  complexing  compound  discovered  in  Lithophaga  by  Jaccarini, 
Bannister,  and  Micallef  (1968).  There  is,  however,  one  significant  difference  between  the  Jurassic  and 
recent  species.  The  Jurassic  species  had  no  posterior  gape,  while  the  modern  species  does.  H.  gallicana 
may  frequently  start  its  byssal  life  attached  in  the  opening  of  an  annelid  boring  (Parfitt  1871),  which  is 
then  enlarged  and  deepened  into  the  substrate.  The  second  species,  H.  arctica  (Linne)  (text-fig.  4a)  is  a 
byssal  nestler  which  is  not  normally  associated  with  borings,  but  which  may  fortuitously  occur  there. 
It  is  commonly  found  single  in  association  with  masses  of  byssate  bivalves  like  Mytilus,  although 
Ockelmann  (1958)  records  it  occurring  as  monotypic  clusters  in  Greenland.  This  species  is  similar  to 
the  non-boring  Jurassic  forms  described  above,  but  is  slightly  more  elongate  and  the  posterior  carinae 
have  more  lamellose  tuberculate  ornament.  H.  arctica  and  H.  gallicana  are  readily  distinguished  in 
the  larval  stage,  but  the  adult  morphologies  intergrade  because  of  overlap  in  habitat,  and  it  is  not 
always  possible  to  separate  them  perfectly  on  features  of  the  hard-part  anatomy.  There  is  therefore 
little  reason  to  attempt  to  separate  the  Yolgian  ecomorphs  into  different  species. 

Bivalve  borings  in  phosphatized hardgrounds.  Although  bivalves  are  commonly  found  associated  with 
calcareous  substrates,  there  appear  to  be  relatively  few  recorded  examples  of  them  penetrating 
phosphatized  hardgrounds.  It  is  clear  that  the  Spilsby  nodules  were  already  phosphatized  at  the  time 
of  attack;  any  doubts  that  could  be  raised  can  be  dispelled  by  the  occurrence  of  the  borings  into  fossil 
bone,  which  is  a primary  phosphate.  Carcelles  (1944)  recorded  Lithophaga  ( Diberus ) penetrating  the 
plates  of  Glyptodon  and  Boreske,  Goldberg,  and  Cameron  (1972)  reported  the  occurrence  of  Miocene 
bivalve  borings  in  the  bone  of  Squalodon  and  attributed  them  to  Parapholas.  They  also  recorded  the 
occurrence  of  such  borings  in  mammoth  tusks.  If  these  borings  in  phosphatized  substrates  are 
constructed  by  mechanical  processes  there  are  no  problems.  However,  if  chemical  techniques  are  to 
be  invoked,  further  research  along  the  lines  of  Jaccarini  et  al.  (1968)  should  be  investigated. 

Environment  of  deposition  of  Basal  Spilsby  Nodule  Bed.  The  Basal  Spilsby  Nodule  Bed  formed  a 
hardground  not  of  a continuous  type  (e.g.  type  2 of  Goldring  and  Kazmierczak  (1974,  p.  957)),  but  of 
isolated  nodules  surrounded  by  glauconitic  silty  matrix.  The  nodules  show  a complex  depositional 
history  and  correspond  partly  to  the  hiatus  concretions  of  Voigt  (1968),  whose  observations  were 
based  on  Liassic  calcareous  concretions.  These  calcareous  concretions  were  formed  by  coalescence  of 
concretions  of  different  age.  The  younger  concretions  envelope  the  older,  the  concretions  themselves 
being  of  early  diagenetic  origin.  Voigt  recognized  the  following  cyclic  sequence  of  events:  1, 
formation  of  concretion;  2,  washout;  3,  corrosion,  boring,  and  encrustation;  4,  burial.  The 
Cenomanian  phosphatized  nodules  described  by  Kennedy  and  Garrison  (1975)  correspond  more 
closely  to  the  Spilsby  nodules.  For  discussion  of  earlier  studies  on  phosphatized  horizons  of 
condensation  see  Bruckner  (1977).  Kennedy  and  Garrison  (1975)  propose  the  following  sequence  for 
the  formation  of  nodules  that  are  largely  composed  of  fossil  moulds:  1,  infilling  of  shell  by  sediment; 
2,  burial;  3,  mould  cementation  (probably  by  high-magnesian  calcite);  4,  dissolution  of  aragonitic 
shell;  5,  disinterment;  and  6,  phosphatization,  boring,  and  encrusting.  The  Basal  Spilsby  Nodules 
appear  to  have  formed  under  similar  conditions,  although  it  is  believed  here  that  phosphatization 
probably  took  place  at  depth  in  the  sediment  and  not  on  the  surface  of  the  sea  floor  as  Kennedy  and 
Garrison  (1975,  p.  357)  suggest.  It  is  not  possible  to  see  deep  burrowing  bivalves  like  Pleuromya  and 


778 


PALAEONTOLOGY,  VOLUME  23 


Lucina  in  life  position  in  the  Basal  Spilsby  Nodule  Bed,  although  they  are  particularly  common  as 
heavily  darkened  phosphatized  internal  moulds.  However,  in  the  Speeton  Clay  of  the  Yorkshire  coast 
(Lower  Cretaceous),  deep  burrowers  such  as  Thracia  and  Pleuromya  are  commonly  preserved  in  life 
position  as  weakly  phosphatized,  pink  or  pale-brown  internal  moulds  with  some  original  shell 
attached.  These  have  clearly  never  been  exposed  on  the  sea  floor;  those  that  have  become  exposed  and 
occur  in  the  reworked  nodule  beds  are  usually  blackened  on  the  exterior  and  may  show  signs  of 
erosion. 


Nl  U 


diagenesis 


i-  ■ . • ■ 
* « . 

T-  - 


o 


o 


rm 


5.  Sedimentation  & 6.  Second  Phosphatisation  7 Exhumation  & 

diagenesis  reworking 


8.  Boring 


9.  Sedimentation,  lO.  Exhumation  & 11.  Boring  12.  Final  burial  & 

diagenesis  & reworking  phosphatisation 

third  phosphatisation 


text-fig.  6.  Simplified  diagrammatic  representation  of  the  sequence  of  events  leading  to  the  formation  of  the 

Basal  Spilsby  Nodule  Bed. 


KELLY:  JURASSIC  BIVALVE  SQUATTER? 


779 


The  preservation  of  many  fossils  in  the  Basal  Spilsby  Nodule  Bed  as  phosphatized  internal  moulds 
suggests  that  the  confining  shell  walls  have  provided  a reduced  zone  within  the  sediment.  The 
phosphatization  occurred  within  this  zone  and  appears  first  in  deep  recesses  such  as  the  umbonal 
infilling  in  bivalves,  and  may  appear  weaker  towards  the  commissure,  especially  so  in  forms  with 
commissural  gape